<p>Tokom poslednje decenije efikasnost DC/DC konvertora, brzina odziva i snaga<br />koju mogu da isporuče ka potrošaču postali su značajan faktor prilikom<br />konstruisanja i izbora ovih kola. U mnogim istraživanjima je potvrđeno da<br />korišćenje višefaznih topologija, uzajamno sprezanje faza i simetrija faza<br />unutar kola igraju značajnu ulogu u unapređenju navedenih karakteristika DC/DC konvertora.<br />U ovoj disertaciji je predstavljen novi simetrični dizajn višefaznih<br />spregnutih induktora kod kojih svi namotaji dele jedno, zajedničko, feritno<br />jezgro. Projektovane strukture su analizirane u laboratorijskim uslovima u<br />frenkvencijskom opsegu od 50 kHz do 40 MHz, a merni rezultati su potvrđeni<br />simulacijama u specijalizovanom softveru. Uzorci su testirani u radnim<br />uslovima pomoću šestofaznog DC/DC buck konvertora. Rezultati testova su<br />pokazali da primena konstruisanih induktorskih struktura dovodi do<br />poboljšanja brzine odziva i porasta efikasnosti test kola.</p> / <p>In the last decade efficiency of DC/DC converters, transient response and<br />delivered power, become important factor in development and selecting<br />these circuits. Numerous researches confirm benefits of using multiphase<br />topologies, mutual coupling of phases and symmetry of phases in the circuit<br />of DC/DC converters.<br />In this dissertation, new symmetric designs of multiphase coupled inductors<br />placed in the same core have been presented. Designed structures have<br />been tested in laboratory conditions in frequency range from 50 kHz to 40<br />MHz, and results have been confirmed by simulations in specialized<br />software. Testing of samples in operating conditions have been performed<br />with six-phase buck converter circuit. These experiments confirm that<br />developed inductors improve transient response and efficiency of test circuit.</p>

<div>Vacuum is needed in many applications and, there are many types of pumps that can provide the vacuum level needed. One widely used pump is the liquid-ring vacuum pump, which does not involve any solid-solid contacts at interfaces where moving and stationary parts meet. Though liquid-ring vacuum pumps are efficient and robust, manufacturers have aggressive goals on improving efficiency, performance, and range of operations.</div><div> </div><div> In this research, time-accurate computational fluid dynamic (CFD) analyses were performed to study the flow mechanisms in a liquid-ring vacuum pump to understand how it works and how the design can be improved. Based on the understanding gained, a physics based reduced order model was developed for preliminary design of the liquid ring vacuum pumps.</div><div> </div><div> In the CFD analyses, the liquid (water) was modeled as incompressible, the gas (air) as an ideal gas, and turbulence by the shear-stress transport model. The gas-liquid interface was resolved by using the volume-of-fluid method, and rotation of the impeller was enabled by using a sliding mesh. Parameters examined include the suction pressure (75, 300, and 600 Torr) and the impeller's rotational speed (1150, 1450 and 1750 rpm) with the temperature of the gas at the inlet of the suction chamber kept at 300 K and the pressure at the outlet of the exhaust chamber kept at one atmosphere. The CFD solutions generated were verified via a grid sensitivity study and validated by comparing with experimental data. When compared with experiments, results obtained for the flow rate of the gas ingested by the pump had relative errors less than 6\% and results obtained for the power consumed by the pump had relative errors less than 13\%.</div><div> </div><div> Results obtained show the shape of the liquid ring to play a dominant role in creating the expansion ratio or the vacuum needed to draw air into the pump through the suction port and the compression ratio needed to expel the air through the discharge ports. Results were generated to show how centrifugal force from rotation and how acceleration/deceleration from the difference in pressure at the pump's inlet and outlet along with the eccentricity of the impeller relative to the pump's housing affect the shape of the liquid ring. Results were also generated to show how the rotational speed of the impeller and the pressure at the suction port affect the nature of the gas and liquid flow in the pump and the pump’s effectiveness in creating a vacuum. </div><div> </div><div> With the knowledge gained from the CFD study, a physics-based reduced-order model was developed to predict air ingested and power consumed by the pump as well as the liquid ring shape and pressure of the gas and liquid in the pump as a function of design and operating parameters. This model was developed by recognising and demonstrating that the amount of air ingested and power consumed by the pump is strongly dependent on the shape and location of the liquid ring surface. The flow rates of the gas ingested by the pump and the power consumed by the pump predicted by the model were compared with experimental data and relative errors were less than 12\% and 17\% respectively.</div>

This thesis presents the simulation, design, and hardware implementation of a modified multiphase boost converter. Converter design must consider noise imposed on input and output nodes which connect to and influence the operation of other devices. Excessive noise introduces EMI which can damage sensitive circuits or impede their operation. High ripple current degrades battery lifetime and reduces operating efficiency in connected systems such as PV arrays. Converters with high ripple current also experience greater peak conduction loss and require larger components. A two-phase implementation of a modified boost converter demonstrates the input current filtering benefits of the modified topology with increased power capacity. In a 12V to 19V 95W design, the modified multiphase design exhibits a reduced input current ripple of 1.103% compared to the 9.096% of the standard multiphase design while imposing minimal detriment to overall converter efficiency. The modified topology uses two inductors and one feedback capacitance per phase. Larger value inductors generally exhibit lower current ratings as well as larger size. The split inductance of the modified multiphase topology can be designed for occupation of less total volume than the single inductance of the standard multiphase topology.

<p>Multiphase fields find wide applications in the fields of combustion, sprays, turbomachinery, heating and cooling systems, blasts, energetic materials, and several more areas of engineering interest. As the efficiency and performance of these systems depend heavily on the underlying multiphase field, studying their intricate structural features becomes important. The current study follows the development of a three-dimensional Wide-Angle Relay Plenoptic (WARP) imaging system with two image quadruplers for the tomographic imaging of multiphase fields. 3D printed targets were used to simulate both semi-transparent as well as opaque particle fields to emulate multiphase systems. Tomographic reconstruction of the targets was performed using the iterative MART reconstruction algorithm in a commercial image processing software. Reconstructions were performed at different angular separations between the cameras as well as for varied separation distance between the object and the imaging system. Quantitative analysis of the reconstruction quality of the developed system was performed to study the effectiveness and accuracy of this system in imaging multiphase fields. The effect of varying different system parameters on reconstruction quality has been studied to evaluate the best system configuration for imaging multiphase fields.</p>

Modelling multiphase flow, more specifically particle-laden flow, poses multiple challenges. These difficulties are heightened when the particles are differentiated by a set of “internal” variables, such as size or temperature. Traditional treatments of such flows can be classified in two main categories, Lagrangian and Eulerian methods. The former approaches are highly accurate but can also lead to extremely expensive computations and challenges to load balancing on parallel machines. In contrast, the Eulerian models offer the promise of less expensive computations but often introduce modelling artifacts and can become more complicated and expensive when a large number of internal variables are treated. Recently, a new model was proposed to treat such situations. It extends the ten-moment Gaussian model for viscous gases to the treatment of a dilute particle phase with an arbitrary number of internal variables. In its initial application, the only internal variable chosen for the particle phase was the particle diameter. This new polydisperse Gaussian model (PGM) comprises 15 equations, has an eigensystem that can be expressed in closed form and also possesses a convex entropy. Previously, this model has been tested in one dimension. The PGM was developed with the detonation of radiological dispersal devices (RDD) as an immediate application. The detonation of RDDs poses many numerical challenges, namely the wide range of spatial and temporal scales as well as the high computational costs to accurately resolve solutions. In order to address these issues, the goal of this current project is to develop a block-based adaptive mesh refinement (AMR) implementation that can be used in conjunction with a parallel computer. Another goal of this project is to obtain the first three-dimensional results for the PGM. In this thesis, the kinetic theory of gases underlying the development of the PGM is studied. Different numerical schemes and adaptive mesh refinement methods are described. The new block-based adaptive mesh refinement algorithm is presented. Finally, results for different flow problems using the new AMR algorithm are shown, as well as the first three-dimensional results for the PGM.

Fuel leakage through a fill limit vent valve (FLVV) inside a fuel tank is an important factor to consider during the design of a new tank. The performance of the carbon canister which absorbs the hydrocarbon can be compromised if fuel manages to escape through the valve, so called Liquid Carry Over (LCO) and thus not fulfilling the fuel emission requirements. As of today this is not thoroughly investigated using experiments nor Computational Fluid Dynamics. The main focus of this study was to develop a method to simulate the behaviour of the FLVV during various driving conditions at an early design stage and if this gives rise to fuel escaping through the FLVV. This method was later to be validated with an experimental set-up and later used to perform some simulations to investigate LCO by varying different parameters such as fuel level and different types of driving. What happens when the canister is purging was also investigated to see if it has a pronounced effect on LCO. Purging is when hydrocarbons, absorbed by the canister, are sent to the engine and giving rise to an under pressure in the tank.The method was developed to run on a cluster utilizing 200 Central Processing Unit Cores where each simulated physical second required an average of 3 hours of simulation time.The flow inside the tank was simulated using a Volume Of Fluid (VOF) multiphase model and the dynamic behaviour of the floater inside the FLVV was simulated using an overset mesh with a Dynamic Fluid Body Interaction.The movement of the simulated dynamic floater was validated with an experimental set-up specifically developed for the overset mesh validation and the motion of the floater was captured at a fairly accurate level.A prototype for an experimental tank was also developed and produced to validate the VOF set-up used for sloshing inside the tank which was utilized on the real tank but due to time limitation the experiments were not performed. The results from the parameter investigation showed that LCO was present in cases with high fuel level inside the tank 95 % and that an aggressive driving gives rise to a higher level of LCO compared to normal driving. Simulations with a fuel level of 85 % and lower showed no evidence of LCO for this particular tank model. The purging of the tank induced a pumping effect giving rise to a higher level of LCO pumped through by the floater.

This Master Thesis presents TMVOC simulations of a NAPL-spill (non-aqueous phase liquid) and following pumping remediation. TMVOC is a simulation program for three-phase non-isothermal multicomponent flow in saturated-unsaturated heterogeneous media. The models presented are based on an actual remediation project. The aim of the thesis was to study if the historical development of the NAPL-spill could be simulated and how long time the pumping remediation would take. A 3D-model and a radially symmetric cylindrical model were created. A large effort of the work done was in taking the complex TMVOC model in use and modifying it for the problem at hand. Therefore, the numerical results of the simulations should be considered as preliminary and as forming basis for future studies. The results from the spill simulation and historical pumping simulation indicated that the spill volume could be less than the estimated 1400 m3, perhaps around 700 m3, assuming a leakage time of 30 years. The historical pumping simulation of a 700 m3 diesel spill showed good agreement with measured values for some wells, but overestimated the recovery in other wells. The overestimation could be due to the fact that the 3D-model did not take seasonal changes in the groundwater level into consideration. Also, the model did not account for any heterogeneity or compartmentalization in soil material properties that could explain the differences between the wells.  Assuming the same spill of 700 m3, future pumping was simulated. The results from these simulations indicated the remediation time to be long due to fast decreasing mobility of the NAPL phase. The NAPL flow rate to the wells was halved in a couple of years. Much of the NAPL was distributed over a large area at near residual saturation with the highest NAPL saturation found at the opposite side of the pumping wells in the model.   Future simulation studies should address the effect of discretization as well as the effect of uncertainties in material properties e.g. conductivity, residual NAPL saturation and soil heterogeneity.

Studying the interactions of solid particles and deformable gas<br />bubbles in viscous liquids is very important in many applications,<br />especially in mining and chemical industries. These interactions<br />involve liquid-solid-air multiphase flows and an<br />arbitrary-Lagrangian-Eulerican (ALE) approach is used for the direct<br />numerical simulations. In the system of rigid particles and<br />deformable gas bubbles suspended in viscous liquids, the<br />Navier-Stokes equations coupled with the equations of motion of the<br />particles and deformable bubbles are solved in a finite-element<br />framework. A moving, unstructured, triangular mesh tracks the<br />deformation of the bubble and free surface with adaptive refinement.<br />In this dissertation, we study four problems. In the first three<br />problems the flow is assumed to be axisymmetric and two dimensional<br />(2D) in the fourth problem.<br /><br />Firstly, we study the interaction between a rising deformable bubble<br />and a solid wall in highly viscous liquids. The mechanism of the<br />bubble deformation as it interacts with the wall is described in<br />terms of two nondimensional groups, namely the Morton number (Mo)<br />and Bond number (Bo). The film drainage process is also<br />considered. It is found that three modes of bubble-rigid wall<br />interaction exist as Bo changes at a moderate Mo.<br />The first mode prevails at small Bo where the bubble deformation<br />is small. For this mode, the bubble is<br /> hard to break up and will bounce back and eventually attach<br />to the rigid wall. In the second mode, the bubble may break up after<br />it collides with the rigid wall, which is determined by the film<br />drainage. In the third mode, which prevails at high Bo, the bubble<br />breaks up due to the bottom surface catches up the top surface<br />during the interaction.<br /><br />Secondly, we simulate the interaction between a rigid particle and a<br />free surface. In order to isolate the effects of viscous drag and<br />particle inertia, the gravitational force is neglected and the<br />particle gains its impact velocity by an external accelerating<br />force. The process of a rigid particle impacting a free surface and<br />then rebounding is simulated. Simplified theoretical models are<br />provided to illustrate the relationship between the particle<br />velocity and the time variation of film thickness between the<br />particle and free surface. Two film thicknesses are defined. The<br />first is the thickness achieved when the particle reaches its<br />highest position. The second is the thickness when the particle<br />falls to its lowest position. The smaller of these two thicknesses<br />is termed the minimum film thickness and its variation with the<br />impact velocity has been determined. We find that the interactions<br />between the free surface and rigid particle can be divided into<br />three regimes according to the trend of the first film thickness.<br />The three regimes are viscous regime, inertial regime and jetting<br />regime. In viscous regime, the first film thickness decreases as the<br />impact velocity increases. Then it rises slightly in the inertial<br />regime because the effect of liquid inertia becomes larger as the<br />impact velocity increases. Finally, the film thickness decreases<br />again due to Plateau-Rayleigh instability in the jetting regime.<br />We also find that the minimum film thickness corresponds to an<br />impact velocity on the demarcation point between the viscous and<br />inertial regimes. This fact is caused by the balance of viscous<br />drag, surface deformation and liquid inertia.<br /><br />Thirdly, we consider the interaction between a rigid particle and a<br />deformable bubble. Two typical cases are simulated: (1) Collision of<br />a rigid particle with a gas bubble in water in the absence of<br />gravity, and (2) Collision of a buoyancy-driven rising bubble with a<br />falling particle in highly viscous liquids. We also compare our<br />simulation results with available experimental data. Good agreement<br />is obtained for the force on the particle and the shape of the<br />bubble.<br /><br />Finally, we investigated the collisions of groups of bubbles and<br />particles in two dimensions. A preliminary example of the oblique<br />collision between a single particle and a single bubble is conducted<br />by giving the particle a constant acceleration. Then, to investigate<br />the possibility of particles attaching to bubbles, the interactions<br />between a group of 22 particles and rising bubbles are studied. Due<br />to the fluid motion, the particles involved in central collisions<br />with bubbles have higher possibilities to attach to the bubble. / Ph. D.

This project aimed to investigate the capability of the STAR CCM+ software when predicting the flow with particles using Lagrangian Particle Tracking and Discrete Element Method. The first part pertained to rectangular channel flow, with ratio between height of the channel and particle diameter (2h/Dp ) of 15. It was found out that simulations of particles in a channel come with many diculties. Such as, obtaining accurate pressure drop results using DEM when comparing to DNS simulations including particles within a reasonable computational time. The second part consisted of a simulation of the off-centred mixing tank. As the use of DEM caused numerical issues, another modeling approach was used. Therefore, the Lagrangian Particle Tracking was used. The outcome of the project is the sensitivity study of the forces which can be applied to the particles. The finding was that the Shear Lift force and the Virtual Mass force have a negligible contribution in regards to the particles distribution. In addition to this, it was also discovered that the turbulence model has a large effect on the particles in the near-wall region. Choosing an isotropic turbulence model resulted in clustering of the particles near the wall, therefore an anisotorpic turbulence model needed to be used.

Multiphase flows are ubiquitous across engineering disciplines: water-sediment river flows in civil engineering, oil-water-sand transportation flows in petroleum engineering; and sorbent-flue gas reactor flows in chemical engineering. These multiphase flows can include a combination of momentum, heat, and mass transfer. Studying and understanding the behavior of multiphase, multiphysics flow configurations can be crucial for safe and efficient engineering design. In this work, a framework for the development and validation, verification and uncertainty quantification (VVUQ) of subgrid models for heat transfer in multiphase flows is presented. The framework is developed for a carbon capture reactor; however, the concepts and methods described in this dissertation can be generalized and applied broadly to multiphase/multiphysics problems. When combined with VVUQ methods, these tools can provide accurate results at many length scales, enabling large upscaling problems to be simulated accurately and with calculable errors. The system of interest is a post-combustion solid-sorbent carbon capture reactor featuring a solid-sorbent bed that is fluidized with post-combustion flue gas. As the flue gas passes through the bed, the carbon dioxide is exothermically adsorbed onto the sorbent particle’s surface, and the clean gas is passed onto further processes. To prevent overheating and degradation of the sorbent material, cooling cylinders are immersed in the flow to regulate temperatures. Simulating a full-scale, gas-particle reactor using traditional methods is computationally intractable due to the long time scale and variations in length scales: reactor, O(10 m); cylinders, O(1 cm); and sorbent particles, O(100 um). This research developed an efficient subgrid method for simulating such a system. A constitutive model was derived to predict the effective suspension-cylinder Nusselt number based on the local flow and material properties and the cylinder geometry, analogous to single-phase Nusselt number correlations. This model was implemented in an open source computational fluid dynamics code, MFIX, and has undergone VVUQ. Verification and validation showed great agreement with comparable highly-resolved simulations, achieving speedups of up to 10,000 times faster. Our model is currently being used to simulate a 1 MW, solid-sorbent carbon capture unit and is outperforming previous methods in both speed and physically accuracy. / 2017-06-21T00:00:00Z