Submitted by Maria Medeiros (maria.dilva1@ufcg.edu.br) on 2018-08-16T12:27:26Z No. of bitstreams: 1 AMANDA SANTANA PEITER - TESE (PPGEQ) 2017.pdf: 4504888 bytes, checksum: 524c2d78079fab537fef86ee47f99f28 (MD5) / Made available in DSpace on 2018-08-16T12:27:26Z (GMT). No. of bitstreams: 1 AMANDA SANTANA PEITER - TESE (PPGEQ) 2017.pdf: 4504888 bytes, checksum: 524c2d78079fab537fef86ee47f99f28 (MD5) Previous issue date: 2017-09-15 / O consumo de combustíveis derivados do petróleo aumenta a cada ano, o que resulta no aumento da poluição atmosférica. Devido a isso a economia mundial necessita de fontes alternativas de energia. O biodiesel existe como alternativa ao diesel. Biodiesel pode ser obtido a partir da reação de transesterificação, na qual os triglicerídeos presentes nos óleos vegetais ou gorduras animais reagem com álcool de cadeia curta na presença de um catalisador para obter os ésteres de ácidos graxos. Neste trabalho foi encontrada a cinética da reação de transesterificação de óleo de soja com metanol e com etanol, na presença de hidróxido de sódio, para produzir biodiesel. Para o biodiesel metílico a cinética foi de primeira ordem e para o etílico a cinética foi de segunda ordem. Esses dados da cinética foram utilizados para fazer simulações das reações. A técnica de fluidodinâmica computacional CFD permite um estudo profundo e detalhado, inclusive análises de campos de velocidade e pressão, sem a necessidade de técnicas experimentais avançadas de análise ou custos financeiros para montagem de planta piloto. Essa técnica é desenvolvida principalmente em pacotes comerciais como o Ansys, que oferece diversas interfaces gráficas e modelos. Neste trabalho foi criada a geometria e a malha para o tanque com os diferentes impelidores e inseridas as condições da reação para encontrar os resultados. O modelo utilizado foi o de turbulência k- ε, a cinética foi considerada de primeira ordem para o metanol e de segunda ordem para o etanol e o sistema considerado monofásico e multicomponente. A simulação para o etanol não se mostrou satisfatória, pois as frações mássicas simuladas não tiveram o comportamento semelhante ao experimental. Já para as reações com metanol, as frações mássicas obtidas pela simulação ficaram bem próximas as encontradas experimentalmente, mostrando que a simulação foi satisfatória. / The consumption of fuels derived from petroleum increases every year, resulting in increased air pollution. Because of this the world economy needs alternative sources of energy. Biodiesel exists as an alternative to diesel. Biodiesel can be obtained from the transesterification reaction in which the triglycerides present in vegetable oils or animal fats react with short chain alcohol in the presence of a catalyst to obtain the fatty acid esters. In this work the kinetics of the reaction of transesterification of soybean oil with methanol and with ethanol in the presence of sodium hydroxide to produce biodiesel was found. For the methyl biodiesel the kinetics was of first order and for the ethyl the kinetics was of second order. These kinetic data were used to simulate the reactions. The CFD computational fluid dynamics technique allows in-depth and detailed study, including velocity and pressure field analyzes, without the need for advanced experimental analysis techniques or financial costs for pilot plant assembly. This technique is mainly developed in commercial packages such as Ansys, which offers several graphical interfaces and models. In this work the geometry and the mesh were created for the tank with the different impellers and the conditions of the reaction were inserted to find the results. The k-ε turbulence model was used, kinetics was considered first-order for methanol and second-order for ethanol and the system considered single-phase and multicomponent. The simulation for the ethanol was not satisfactory, since the simulated mass fractions did not have the behavior similar to the experimental one. As for the reactions with methanol, the mass fractions obtained by the simulation were very close to those found experimentally, showing that the simulation was satisfactory.

Dans cette thèse, une formulation pour l’extraction du couple « champ lointain » d’un rotor d’hélicoptère en vol stationnaire est présentée. Cette formulation est dérivée de la méthode d’extraction de la traînée « champ lointain » d’un avion, basée sur les travaux de van der Vooren et Destarac [?, ?, ?]. Un outil développé à l’Onera à partir de cette théorie permet de donner une analyse complète de la traînée aérodynamique d’un avion. Il est basé sur l’analyse physique et locale de l’écoulement calculé autour de l’aéronef, et décompose la traînée totale, aussi appelée traînée mécanique, en composantes physiques. Ces composantes physiques peuvent être définies comme suit : 1) la traînée d’onde, 2) la traînée visqueuse, 3) la traînée induite. L’adaptation de la méthode d’extraction de la traînée d’un avion à un rotor en vol stationnaire nécessite l’utilisation du couple rotor à la place de la traînée de l’avion, ce qui donne la décomposition suivante : 1) le couple d’onde, 2) le couple visqueux, 3) le couple induit. Les simulations de rotor diffèrent de celles de l’avion dans la mesure où les équations d’Euler (ou RANS) ne sont pas écrites dans le même repère de référence : les simulations d’avion utilisent généralement une formulation en vitesse relative tandis que les simulations d’un rotor d’hélicoptère utilisent la vitesse absolue projetée dans le repère relatif. Cette différence conduit à deux formulations différentes des équations de l’écoulement, et nécessairement deux formulations différentes de l’extraction de la traînée ou du couple. Ce changement de repère implique aussi des changements dans les quantités thermodynamiques utilisées, en particulier l’utilisation de la rothalpie à la place de l’enthalpie d’arrêt pour déterminer le couple dû aux phénomènes irréversibles. Une application de cette méthode est présentée sur un rotor quadripale créé pour cette étude et montre comment cette nouvelle approche peut améliorer la précision de l’extraction des performances d’un rotor à partir de résultat issu de la simulation numérique. / In this thesis, a formulation for “far-field” torque extraction in the case of a hovering rotor is presented. This formulation is derived from an aircraft “far-field” drag extraction method, based on van der Vooren and Destarac’s works [?, ?, ?]. A tool was previously developed at Onera to give an aerodynamic comprehensive analysis of aircraft drag, based on physical and local analysis of the computed flow field surrounding the aircraft. It decomposes the total drag, also called mechanical drag, into its physical components. These physical components can be defined as : 1) wave drag, 2) viscous drag, 3) induced drag. The adaptation of the method to a rotor in hover leads to consider rotor torque instead of aircraft drag, which gives the following decomposition : 1) wave torque, 2) viscous torque, 3) induced torque. Rotor simulations differ from aircraft ones inasmuch as the Euler (or RANS) equations are not written in the same reference frame : aircraft simulations use the relative velocity formulation while rotor simulations use the absolute velocity projected in the relative frame. This difference leads to two different formulations of the flow equations, and necessarily two different formulations of the drag or torque extraction. This change of reference frame also implies some changes in the thermodynamical quantities used, in particular the use of the rothalpy instead of the stagnation enthalpy to determine the torque due to irreversible phenomena. An application of this method is described on a four-bladed rotor created for this study and shows how this method can improve rotor performance extraction from numerical simulations.

The purpose of this master thesis is to investigate the usage of the Lattice Boltzmann Method (LBM) of Computational Fluid Dynamics (CFD) for real-time prediction of indoor air flows inside a data center module. Thermal prediction is useful in data centers for evaluating the placement of heat-generating equipment and air conditioning. To perform the simulation a program called RAFSINE was used, written by Nicholas Delbosc at the University of Leeds, which implemented LBM on Graphics Processing Units (GPUs) using NVIDIA CUDA. The program used the LBM model called Bhatnagar-Gross-Krook (BGK) on a 3D lattice and had the capability of executing thermal simulations in real-time or faster than real-time. This fast rate of execution means a future application for this simulation could be as a predictive input for automated air conditioning control systems, or for fast generation of training data sets for automatic fault detection systems using machine learning. In order to use the LBM CFD program even from hardware not equipped with NVIDIA GPUs it was deployed on a remote networked server accessed through Virtual Network Computing (VNC). Since RAFSINE featured interactive OpenGL based 3D visualization of thermal evolution, accessing it through VNC required use of the VirtualGL toolkit which allowed fast streaming of visualization data over the network. A simulation model was developed describing the geometry, temperatures and air flows of an experimental data center module at RISE SICS North in Luleå, Sweden, based on measurements and equipment specifications. It was then validated by comparing it with temperatures recorded from sensors mounted in the data center. The thermal prediction was found to be accurate on a room-level within ±1° C when measured as the average temperature of the air returning to the cooling units, with a maximum error of ±2° C on an individual basis. Accuracy at the front of the server racks varied depending on the height above the floor, with the lowest points having an average accuracy of ±1° C, while the middle and topmost points had an accuracy of ±2° C and ±4° C respectively. While the model had a higher error rate than the ±0.5° C accuracy of the experimental measurements, further improvements could allow it to be used as a testing ground for air conditioning control or automatic fault detection systems.

This thesis has developed wetting characteristics of coal dust from Herrin coal seam in Illinois for improved respirable coal and quartz dust (< 10µm) control. Wetting characteristics were assessed by a wettability technique called Fixed Time Wettability (FTW) method which simulates the actual underground wetting environment. Immediate roof and immediate floor materials bulk samples were found to be 100% wettable by water alone, whereas coal seam samples were 55-90% wettable. Two variables: contact time between water droplets and dust particles and, the temperature of the water were found to affect the degree of wettability. FTW was carried out at as a function of water temperature (7.2oC to 40.5oC) and contact time (10 to 25 seconds) which increased wettability 2-18% and 8-32%, respectively. In addition, the wetted and un-wetted samples collected from wettability experiments were subjected to Scanning Electron Microscopy (SEM) and particle size distribution analyses. It was found that particles are irregular in shape in both wetted and un-wetted fractions and some agglomeration of fine particles was observed in a few un-wetted fraction samples. The concept of increased contact time was incorporated in engineering controls by increasing the wet scrubber filter screen inclination (which is a function of the screen surface area). The filter screen inclination was minimizing the pressure drop across it using Computational Fluid Dynamics (CFD) analysis. It was found that 60 degree (with respect to the vertical) would provide minimal pressure drop and should be considered for implementation. Additional research should be conducted on assessing selective wetting of different particle sizes of coal and quartz in airborne dust, developing data on wettability as a function of particle size and implementing the developed concepts such as contact time and temperature in the field for improved dust control.

Nos últimos anos, a utilização da energia eólica vem apresentando uma tendência de aumento. Um dos principais aspectos para determinar a viabilidade técnica e econômica de uma instalação eólica é a avaliação precisa da distribuição das velocidades de vento na área de aproveitamento. A instalação de turbinas eólicas em áreas com terrenos complexos tem determinado a necessidade de aprimorar a metodologia de previsão do campo de velocidades do vento visando à melhor determinação da distribuição dos equipamentos e aproveitamento do potencial existente. Neste contexto, esta dissertação apresenta um estudo sobre a aplicação da Dinâmica dos Fluidos Computacional - CFD para avaliação do potencial eólico e o comportamento do vento sobre um modelo de uma superfície de topografia complexa. Resultados numéricos com diferentes alternativas de modelagem do problema são comparados com dados de um experimento em túnel de vento, visando determinar a metodologia adequada para avaliação do problema proposto. As simulações numéricas do escoamento de ar sobre o terreno são realizadas com o uso do programa ANSYS-Fluent 13.0, que utiliza o método de volumes finitos para a solução das equações de Navier-Stokes com médias de Reynolds (RANS). O estudo é dividido em três casos. No primeiro caso, a rugosidade superficial é negligenciada e o problema de fechamento é contornado com a utilização do modelo de turbulência k ω SST. No segundo caso, a rugosidade superficial é estipulada de acordo com modelo utilizado no ensaio experimental e o problema de fechamento é contornado com a utilização do modelo de turbulência k - ε. No terceiro caso, a rugosidade superficial também é estipulada de acordo com modelo utilizado no ensaio experimental e o problema de fechamento é contornado com a utilização do modelo de turbulência k ω SST. Os resultados das simulações são apresentados de forma que se possa observar o perfil de velocidades adimensional sobre a superfície da geometria para cada caso, para que seja possível verificar o campo de velocidades sobre a superfície em estudo. O resultado das simulações são comparados com dados experimentais obtidos em túnel de vento: verifica-se um comportamento similar nos perfis de velocidade alcançados. Através da análise do campo de velocidades sobre a superfície em estudo, pode-se obter a localização que apresenta o melhor potencial eólico de uma região. Este processo é conhecido como Micrositing. / In recent years the use of wind energy has shown an increasing. A key aspect to determine the technical and economic viability for the wind power plant is the accurate assessment of the distribution of wind speeds in the area of utilization. The installation of wind turbines in areas with complex terrain has determined the necessity of improve the methodology for the prediction of wind velocity field in order to better determine the distribution of equipment and utilization of existing potential. In this context this work presents a study on the application of computational fluid dynamics to evaluate the wind potential and the behavior of the wind on a model of a complex surface topography. Numerical results with different alternatives for modeling the problem are compared with data from an experiment in wind tunnel to determine the appropriate methodology for evaluation of the problem. The numerical simulations of the air flow over the terrain are performed using the ANSYS Fluent 13.0 which uses the finite volumes method for solving the Reynolds Averaged Navier Stokes (RANS) equation. The study is divided in three cases. In the first one, the surface roughness is neglected and the closure problem is solved by k ω SST turbulence model. In the second case, the surface roughness is stipulated according to the model used in the experimental test and the closure problem is solved by k ε turbulence model. In the third case, the surface roughness is stipulated according to the model used in the experimental test and the closure problem is solved by k ω SST turbulence model. The simulation results are presented so that one can observe the dimensionless velocity profile on the surface in each case in order to check the velocity field on the surface under investigation. These results are compared with experimental data obtained in wind tunnel which shows consistency with them. By analyzing the velocity field on the surface, it can be verified that the exact location where there is the best wind potential of a region. This process is called Micrositing.

No description available.

SpinChem AB designs rotating bed reactors (RBR:s) that are used to conduct reactions between solid and liquid phase. In this work a model have been developed that predicts the performance of the S2 RBR. Coupled with simulations the model could be used to improve current and future RBR designs.  The model predicts the reaction rate inside the RBR during a de-colourization process where methylene blue was adsorped onto an adsorbent that was chosen during a screening study. The materials investigated were XAD1600N, MN102, activated carbon and IRN99 where the suitability of a material was judged upon the capacity and reproducibility of the process, and the reaction rate dependency on the mass transport. The DOE software MODDE Pro was used to design a parameter study that produced a function that describes the reaction rate of the process as a function of the RPM of the RBR and the temperature, pH and initial concentration of the methylene blue solution. Using simulations in ANSYS Fluent and modelling the function was then converted to the final model. It was then tested by comparing experimental results with simulations of the de-colourization process where the reaction rate inside the RBR was set using the model. IRN99 was chosen as the adsorbent of the process as its capacity was the highest of the four materials studied and its reaction rate displayed a sufficient mass transport dependence. The parameter study produced a function with a high R2, Q2 and reproducibility (&gt;0.85) but displayed a low model validity (-0.2) as a result of the noise in the experimental data. The final model agrees well with the experimental results in the whole parameter space of experiments conducted. The model is meant to be a general model that could be applied to other RBR sizes. It could therefore be used to investigate the performance of new RBR models before they are built or improve current RBR:s. Other geometries has however not been tested during this work and therefore not much can be said about its accuracy when used in this way. / SpinChem AB designar roterande bäddreaktorer (RBR:s) som används för att genomföra reaktioner mellan fast och flytande fas. I detta arbete har en modell utvecklats som förutspår prestandan hos en S2 RBR. Tillsammans med simuleringar kan modellen användas för att förbättra nuvarande och framtida RBR-modeller. Modellen förutspår reaktionshastigheten inuti RBR:en under en avfärgningsprocess där metylenblått adsorberades på ett adsorbent som valdes under en screeningsstudie. Materialet som undersöktes var XAD1600N, MN102, aktiverat kol och IRN99 där lämpligheten av ett material bedömdes på processens kapacitet och reproducerbarhet och reaktionshastighetens beroende på masstransport. DOE-mjukvaran MODDE Pro användes för att utforma en parameterstudie som gav en funktion som beskriver processens reaktionshastighet som en funktion av RBR:ens RPM och temperaturen, pH och initialkoncentrationen av metylenblålösningen. Genom att använda simuleringar i ANSYS Fluent och modellering kunde funktionen omvandlas till en modell som beskriver reaktionens hastighet inne i RBR:en. Den testades sedan genom att jämföra experimentella resultat med simuleringar av avfärgningsprocessen där reaktionshastigheten inuti RBR:en sattes med hjälp av modellen. IRN99 valdes som adsorbent av processen eftersom dess kapacitet var den högsta av de fyra studerade materialen och dess reaktionshastighet visade ett tillräckligt masstransportberoende. Parameterstudien gav en funktion med hög R2, Q2 och reproducerbarhet (&gt; 0,85) men visade på en låg modellgiltighet (-0.2) vilket beror på brus i experimentdata. Den slutliga modellen stämmer väl överens med experimentresultaten i hela parameterrummet för de utförda experimenten. Modellen är avsedd att vara en allmän modell som kan tillämpas på andra RBR-storlekar. Det kan därför användas för att undersöka prestandan hos nya RBR-modeller innan de byggs eller förbättra nuvarande RBR:er. Andra geometrier har dock inte testats under detta arbete och därför kan inte mycket sägas om dess noggrannhet när den används på detta sätt.

Doctor of Philosophy / Biological & Agricultural Engineering / Mark E. Casada / Ronaldo G. Maghirang / Wind erosion is an important problem in many locations, including the Great Plains, that needs to be controlled to protect soil and land resources. This research was conducted to assess the effectiveness of vegetation (specifically, standing vegetation and tree barriers) as controls for wind erosion. Specific objectives were to: (1) measure sand transport and abrasion on artificial standing vegetation, (2) determine porosity and drag of a single row of Osage orange (Maclura pomifera) barrier, (3) assess effectiveness of Osage orange barriers in reducing dust, (4) predict airflow through standing vegetation, and (5) predict airflow and particle collection through Osage orange barriers. Wind tunnel tests were conducted to measure wind speed profiles, relative abrasion energies, and sand discharge rates for bare sand and for two vegetation heights (150 and 220 mm) at various densities of vegetation. Results showed that vegetation density was directly related to threshold velocity and inversely related to sand discharge. The coefficient of abrasion was adversely affected by saltation discharge but did not depend on wind speed. Field tests measured the aerodynamic and optical porosities of Osage orange trees using wind profiles and image analysis, respectively, and an empirical relationship between the two porosities was derived. Vertical wind profiles were also used to estimate drag coefficients. Optical porosity correlated well with the drag coefficient. Field measurements also showed a row of Osage orange barrier resulted in particulate concentration reduction of 15 to 54% for PM2.5 and 23 to 65% for PM10. A computational fluid dynamics (CFD) software (OpenFOAM) was used to predict airflow in a wind tunnel with artificial standing vegetation. Predicted wind speeds differed slightly from the measured values, possibly due to oscillatory motions of the standing vegetation not accounted for in the CFD simulation. OpenFOAM was also used to simulate airflow and particle transport through a row of Osage orange barrier. Predicted and measured wind speeds agreed well. Measured dust concentration reduction at two points (upwind and downwind) were also similar to the predicted results.

abstract: The Volume-of-Fluid method is a popular method for interface tracking in Multiphase applications within Computational Fluid Dynamics. To date there exists several algorithms for reconstruction of a geometric interface surface. Of these are the Finite Difference algorithm, Least Squares Volume-of-Fluid Interface Reconstruction Algorithm, LVIRA, and the Efficient Least Squares Volume-of-Fluid Interface Reconstruction Algorithm, ELVIRA. Along with these geometric interface reconstruction algorithms, there exist several volume-of-fluid transportation algorithms. This paper will discuss two operator-splitting advection algorithms and an unsplit advection algorithm. Using these three interface reconstruction algorithms, and three advection algorithms, a comparison will be drawn to see how different combinations of these algorithms perform with respect to accuracy as well as computational expense. / Dissertation/Thesis / Masters Thesis Mechanical Engineering 2015

abstract: Gnamma pit is an Australian aboriginal term for weathering pit. A mix of weathering and aeolian processes controls the formation of gnamma pits. There is a potential to utilize gnamma as an indicator of paleowind intensity because gnamma growth is promoted by the removal of particles from gnamma pits by wind, a process referred to as deflation. Wind tunnel tests determining the wind velocity threshold of deflation over a range of pit dimensions and particles sizes are conducted. Computational fluid dynamics (CFD) modeling utilizing the Re-Normalisation Group (RNG) K-Epsilon turbulence closure is used to investigate the distribution of wall shear stress and turbulent kinetic energy. An empirical equation is proposed to estimate shear stress as a function of the wind velocity and pit depth dimensions. With this equation and Shields Diagram, the wind velocity threshold for evacuating particles in the pit can be estimated by measuring the pit depth ratio and particle size. It is expected that the pit would continue to grow until this threshold is reached. The wind speed deflation threshold is smaller in the wind tunnel than predicted by the CFD and Shields diagram model. This discrepancy may be explained by the large turbulent kinetic energy in the gnamma pit as predicted by the CFD model as compared to the flat bed experiments used to define the Shields diagram. An empirical regression equation of the wind tunnel data is developed to estimate paleowind maximums. / Dissertation/Thesis / Masters Thesis Geography 2015