When a disordered packed bed, or any heterogeneous media is studied using computational fluid dynamics, the tortuous task of generating a domain and creating a workable mesh presents a challenging issue to Engineers and Scientists. In this Thesis these challenges are addressed in the form of three studies in which both traditional and novel techniques are used to generate packed beds of spheres and cylinders for analysis using computational fluid dynamics, more specifically, the finite volume method. The first study uses a Monte-Carlo method to generate random particle locations for use with a traditional CADbased meshing approach. Computational studies are performed and compared in detail with experimental equivalent beds. In the second study, where there is a need for actual, physical beds to be studied, magnetic-resonance-imaging is used coupled with a novel approach known as image based meshing. In parallel experimental studies are performed on the experimental bed and compared with computational data. In the third study, to overcome fidelity issues with the previous approaches, a physical packed bed is manufactured which is 100% geometrically faithful to its computational counterpart to provide a direct comparison. All three computational studies have shown promising results in comparison with the experimental data described in this Thesis, with the data of Reichelt (1972) and the semi-empirical correlation of Eisfeld & Schnitzlein (2001). All experiments and computational models were carried out by the author unless otherwise stated.

Ingeniero Civil Mecánico / El flujo de burbujas es uno de los tantos ejemplos de los denominados flujos bifásicos, éstos pueden ser encontrados en un gran número de industrias y es simplemente la presencia de más de dos fases en una misma línea de flujo. En particular el flujo de burbujas es usado en el proceso de flotación, este tiene por objetivo la separación de especies minerales mediante la adhesión selectiva de partículas minerales a burbujas de aire, este trabajo pretende empezar con el estudio de este tipo de flujos con la motivación final en las celdas de flotación. El objetivo general del trabajo es utilizar un método numérico para modelar un flujo de burbujas mediante CFD (Computational Fluid Dynamics). El método debe ser aplicado al flujo generado por una corriente de aire impulsada en un estanque con agua, como objetivos específicos debe lograr predecir la forma de la burbuja, el tamaño, la velocidad, trayectoria y la existencia o no de coalescencia, utilizando el método numérico. En la primera parte de este trabajo se realiza una comparación entre modelos teóricos para el volumen, velocidad y trayectoria de la burbuja, con resultados experimentales de algunas bibliografías consultadas, el afán de esto es poder validar un set de modelos teóricos que permitan luego realizar una comparativa con la simulación hecha. Los modelos teóricos encontrados relacionan la forma con diferentes números adimensionales (Eo, Re y M), existe un modelo para el volumen de una burbuja generada a flujo constante, la cual toma en consideración la formación de burbuja en 2 etapas (expansión y desprendimiento) usando como supuesto el radio de la burbuja en la etapa de expansión igual al largo del cuello formado en la etapa de desprendimiento, para la velocidad las múltiples expresiones que existen se relacionan con el diámetro equivalente de la burbuja. Y para la trayectoria, existen variados criterios que difieren unos con otros para determinar una trayectoria recta, zigzag o espiral. El trabajo se centra en un estanque cilíndrico con agua, por el cual se ingresará aire a través de un orificio en su parte inferior, generando así el flujo de burbujas que se desea modelar, el tamaño del orificio y el caudal de aire son variables que se consideran en el estudio preliminar para luego ser determinados según los resultados obtenidos. La simulación realizada, en un entorno 3D, arrojó generación de burbujas con un volumen muy cercano al esperado según los modelos teóricos, frecuencia de formación de burbujas altas, velocidad ascendente de la burbuja por sobre lo esperado, la forma corresponde con la caracterización según los números adimensionales, se observan trayectorias que se pueden clasificar dentro de espirales, además se tiene que el campo de velocidades influye bastante en los distintos resultados obtenidos.

Reverse thrust operations of a model scale Contra-Rotating Open Rotor design were numerically modelled to produce individual rotor thrust and torque results comparable to experimental measurements. The aims of this research were to develop an understanding of the performance and aerodynamics of open rotors during thrust reversal operations and to establish whether numerical modelling with a CFD code can be used as a prediction tool given the highly complex flowfield. A methodology was developed from single rotor simulations initially before building a 3D‘frozen rotor’ steady-state approach to model contra-rotating blade rows in reverse thrust settings. Two different blade pitch combinations were investigated (β1,2 =+30°,- 10° and β1,2 =-10°,-20°). Thrust and torque results compared well to the experimental data and the effects of varying operating parameters, such as rpm and Mach number, were reproduced and in good agreement with the observed experimental behaviour. The main flow feature seen in all the reverse thrust cases modelled, both single rotor and CROR, is a large area of recirculation immediately downstream of the negative pitch rotor(s).This is a result of a large relative pressure drop region generated by the suction surfaces of the negative pitch blades. An initial 3D unsteady sliding-mesh calculation was performed for one CROR reverse thrust case. The thrust and torque values were in poor agreement with experimental values and the disadvantages relating to time costs and required computational resources for this technique were illustrated. However, the results did yield a nominal unsteady variation of thrust and torque due to rotor phase position. Overall the work shows that it may be possible to develop a CROR reverse thrust prediction tool of beneficial quality using CFD models. The research also shows that the frozen rotor approach can be adopted without undermining the fidelity of the results.

This study is aimed at developing modeling methodologies for simulating the flow of air and aerosol particles through fibrous filter media made up of micro- or nano-fibers. The study also deals with modeling particle deposition (due to Brownian diffusion, interception, and inertial impaction) and particle cake formation, on or inside fibrous filters. By computing the air flow field and the trajectory of airborne particles in 3-D virtual geometries that resemble the internal microstructure of fibrous filter media, pressure drop and collection efficiency of micro- or nano-fiber filters are simulated and compared with the available experimental studies. It was demonstrated that the simulations conducted in 3-D disordered fibrous domains, unlike previously reported 2-D cell-model simulations, do not need any empirical correction factors to closely predict experimental observations. This study also reports on the importance of fibers’ cross-sectional shape for filters operating in slip (nano-fiber filters) and no-slip (micro-fiber filters) flow regimes. In particular, it was found that the more streamlined the fiber geometry, the lower the fiber drag caused by a nanofiber relative to that generated by its micron-sized counterpart. This work also presents a methodology for simulating pressure drop and collection efficiency of a filter medium during instantaneous particle loading using the Fluent CFD code, enhanced by using a series of in-house subroutines. These subroutines are developed to allow one to track particles of different sizes, and simulate the formation of 2-D and 3-D dendrite particle deposits in the presence of aerodynamic slip on the surface of the fibers. The deposition of particles on a fiber and the previously deposited particles is made possible by developing additional subroutines, which mark the cells located at the deposition sites and modify their properties to so that they resemble solid or porous particles. Our unsteady-state simulations, in qualitative agreement with the experimental observations reported in the literature, predict the rate of increase of pressure drop and collection efficiency of a filter medium as a function of the mass of the loaded particles.

While most filters are made of pleated fibrous media, almost all existing theories of aerosol filtration are developed for flat media placed perpendicular to the air flow. Expressions developed for flat sheet media do not provide accurate information directly useful for designing a pleated filter, and therefore, most progress made in developing pleated filters is based on empiricism. This study is aimed at establishing an enabling knowledge that allows for a better design and optimization of pleated aerosol filters. This study is focused on developing a predictive simulation method that accounts for the influence of a filter’s micro-scale geometric parameters, such as fiber orientation, as well as its macro-scale features, like pleat shape, in predicting the transient pressure drop and collection efficiency with or without the effects of dust loading. The dual-scale simulation method developed in this work is believed to be the only feasible approach for design and optimization of pleated aerosol filters with the current academic-level computational power. Our study is divided into two major tasks of micro- and macro-scale modeling. Our micro-scale studies are comprised of a series of CFD simulations conducted in virtual 2-D or 3-D fibrous geometries that resemble the internal micro-structure of a fibrous medium. These simulations are intended to isolate the effects of each micro structural parameter and study its influence on the performance of the filter medium. In detail, it is intended to propose a method to predict the performance of micro-structures with fiber size distribution. Also, the effects of micro-structural fiber orientation were investigated. Moreover, we offered methodology to predict the performance of noncircular fibers using available analytical expressions for circular fibers. It is shown that the circumscribed circle for a trilobal shaped fiber gives the best prediction for collection efficiency. In macro-scale simulations, on the other hand, the filter medium is treated as a lumped porous material with its properties obtained via micro-scale simulations. Our results showed that more number of pleats helps better performance of pleated filters, however, if the pleat channel becomes blocked by dust cake then this effect is no longer valid.

Industrial boilers that produce steam or electric power represent a large capital investment as well as a crucial facility for overall plant operations. In real applications, the operation of the superheater for producing high-pressure, high-temperature steam may result in problems frequently caused by ruptured superheater tubes. To make the boiler more efficient, less emission and less prone to tube rupture problems, it is important to understand the combustion and thermal flow behaviors inside the boiler. This study performs a detailed simulation of combustion and thermal flow behaviors inside an industrial boiler. The simulations are conducted using the commercial CFD package FLUENT. The 3-D Navier-Stokes equations and five species transport equations are solved with the eddy-breakup combustion model. Calculation of NOx is performed after obtaining a converged flow, thermal and combustion solution. The results provide insight into the detailed thermal-flow and combustion in the boiler and showing possible reasons for superheater rupture

Gasification is a very efficient method of producing clean synthetic gas (syngas) which can be used as fuel for electric generation or chemical building block for petrochemical industries. This study performs detailed simulations of coal gasification process inside a generic two-stage entrained-flow gasifier to produce syngas carbon monoxide and hydrogen. The simulations are conducted using the commercial Computational Fluid Dynamics (CFD) solver FLUENT. The 3-D Navier-Stokes equations and seven species transport equations are solved with eddy-breakup combustion model. Simulations are conducted to investigate the effects of coal mixture (slurry or dry), oxidant (oxygen-blown or air-blown), wall cooling, coal distribution between the two stages, and the feedstock injection angles on the performance of the gasifier in producing CO and H2. The result indicates that coal-slurry feed is preferred over coal-powder feed to produce hydrogen. On the other hand, coal-powder feed is preferred over coal-slurry feed to produce carbon monoxide. The air-blown operation yields poor fuel conversion efficiency and lowest syngas heating value. The two-stage design gives the flexibility to adjust parameters to achieve desired performance. The horizontal injection design gives better performance compared to upward and downward injection designs.

Calcined coke is the best material for making carbon anodes for smelting of alumina to aluminum. Calcining is an energy intensive industry and a significant amount of heat is wasted in the calcining process. Efficiently managing this energy resource is tied to the profit margin and survivability of a calcining plant. 3-D computational models are developed using FLUENT to simulate the calcining process inside the long slender kiln. Simplified models are employed to simulate the moving petocke bed with a uniform distribution of moisture evaporation, devolatilization, and coke fines entrainment rate with a conjugate radiation-convection-conduction calculation. The results show the 3-D behavior of the flow, the reaction inside the kiln, heat transfer and the effect of the tertiary air on coke bed heat transfer. The ultimate goals are to reduce energy consumption, recover waste-heat, increase thermal efficiency, and increase the product yield.

Unmanned Aerial Vehicle (UAV) is any flying vehicle which is not controlled by actual human pilots sitting in the cockpit but is installed with proper avionics that can either fly autonomously or by using the commands from its base. Some rotorcraft UAVs use a ducted propeller for two main reasons- safety and to increase the thrust produced by the propellers. While ducted rotors can increase the thrust produced, it also adds weight to the UAV. It was therefore hypothesized that by removing part of the duct materials (i.e. adding perforations in the duct) would benefit from both decreased duct weight and increased thrust. However, it is not clear how much trade-off would be between these two factors. Hence, the objective of this study is to explore the relationship between the change of thrust and addition of different numbers or sizes of perforations. Cases with and without duct, and duct with perforations were simulated using a commercial computational fluid dynamic (CFD) software Ansys/Fluent. The physics of the rotating propeller was modeled by a simplified disc with a pressure jump across an infinitesimal volume. Three different RPM speeds of the propellers were simulated by varying the strength of the pressure jump. The results show that the thrust decreases as the duct is added. As perforations are added, the result shows that with more perforations (i.e. more open area on the duct wall), the thrust increases accordingly until the thrust reaches a maximum value without the duct. The result is in contrast to a published experimental data stating that installation of duct can increase thrust. It is speculated that the current duct with a flat wall has caused such difference from the experimental data. Further study is recommended to continue more detailed computational simulation using a duct with cambered airfoil configuration to reduce the aerodynamic losses.

There is an ongoing interest in analyzing the flow characteristics of swimming fish. Biology has resulted in some very efficient motions and formulating these motions is of interest to engineers. One such theory was written by Dr. William Vorus and Dr. Brandon Taravella involving ideal efficiency. It is therefore interesting to test the calculations to see if it is possible to design a motion that can create thrust without necessarily creating vorticity. The computational fluid dynamics software of ANSYS Fluent was used to calculate the resulting flow field of the eel motion to compare with the theoretical values.