Master of Science / Department of Mechanical Engineering / Donald L. Fenton / This project was done to determine the effectiveness of different methods of mitigating the effects of an ammonia release through a pressure relief device in an ammonia refrigeration system. Several methods were considered, and five were selected for further study. The methods chosen for further study were discharge into a tank containing standing water, discharge into the atmosphere, discharge into a flare, discharge into a wet scrubber, and an emergency pressure control system. Discharge into a tank containing standing water is the most common method in existence today but several people in the ammonia refrigeration industry have questioned its reliability. The methods were compared based on a quantitative risk analysis, combining failure rates of each system with ammonia dispersion modeling and the monetized health effects of a system’s failure to contain an ammonia release. It was determined that the release height had a greater influence on the downwind cost impact than any other variable, including weather conditions and release from multiple sources. The discharge into a tank containing standing water was determined to have the lowest failure rate, while the flare system was found to be the most effective in terms of relative overall release consequent cost. The emergency pressure control system is now required by the codes, and any of the other mitigation systems would be very effective when used in conjunction with the emergency pressure control system.

Master of Science / Department of Mechanical and Nuclear Engineering / Kevin Lease / Over the past several years the crack tip opening angle (CTOA) has been identified as one of the key fracture parameters to characterize low constraint stable tearing and instability in structural metallic alloys. This document presents the results of experimental stable tearing characterizations. Characterization methods include optical microscopy and marker band measurements of crack front tunneling. Specific attention is given to the measurement methods used, and also the correlation between CTOA and Delta-5. The effect of tunneling and comparisons with computational results are discussed, and the effect of material and measurement method on CTOA is observed and a clear relationship is seen. Preliminary work on future studies into internal features and behavior is also presented.

Master of Science / Department of Mechanical and Nuclear Engineering / Mohammad H. Hosni / Byron W. Jones / Due to the high occupant density and large number of travelers on commercial aircraft, it is crucial to limit the transport of contaminants and pathogens amongst passengers. In order to minimize the exposure of passengers to various contaminants of different sizes and characteristic, all mechanisms influencing airflow movement within an aircraft cabin need to be understood. The use of personal gaspers on commercial aircraft and their relation to airborne contaminants and pathogens transport is one such mechanism that was investigated. Tracer gas testing using carbon dioxide (CO[subscript]2) was conducted in a wide-body, 11-row Boeing 767 aircraft cabin mockup using actual aircraft components for air distribution. Three separate experiments were conducted investigating the effect of gaspers on the transport of contaminants. The first series of experiments focused on the effect of gaspers on longitudinal transport patterns within an aircraft cabin environment by measuring the concentration of tracer gas along the length of the aircraft cabin. The second experiment investigated what fraction of air a passenger inhales originates from a gasper in relation to the overall cabin ventilation. The final set of experiments determined if gaspers could limit close range person-to-person transmission of exhaled contaminants. Three separate sets of conclusions were drawn, one for each series of experiments. The first conclusion is that gaspers disrupt the longitudinal transport of contaminants within the aircraft cabin. The second conclusion is that less than 5% of the air inhaled by a passenger is originating from a gasper even with a gasper directed at the passenger's face. This low percentage is a result of the turbulent airflow within the aircraft cabin causing the gasper jet to quickly mix with the overall cabin ventilation air. The last conclusion is that gaspers can reduce person-to-person transmission of exhaled contaminants as much as nearly 90% in some cases. In other cases the gaspers are found to have negligible or negative impact on the transmission of contaminants. These conclusions are dependent upon where the tracer gas plume emanated from, the sampling location, and the configuration of gaspers around the tracer gas release point.

Master of Science / Department of Mechanical and Nuclear Engineering / Warren N. White / As traditional resources are becoming scarce, renewable energy is a recent topic receiving greater concern. Among the renewable energies, wind power is a very popular type of energy extracted from wind which is readily available in the environment. The use of wind power all over the world is receiving increased attention. Horizontal axis wind turbines are the most popular equipment for extracting power form the wind. One of the problems of using wind turbines is how to maximize the wind power capture. In this paper, a method for maximizing the rotor power coefficient of a wind turbine is proposed. Simultaneous Perturbation Stochastic Approximation (SPSA) is an efficient way for extremum seeking. It is different from the classical gradient based extremum seeking algorithms. For maximizing the rotor power coefficient, it only needs two objective function measurements to take a step toward the next extremum approximation. The one measurement SPSA is a modification of SPSA method developed in this work. Instead of using measurements of two positions occurring at random directions away from the current position, it uses the measurement of one position in a random direction and the measurement of the current position to estimate the gradient. Usually, the rotor power coefficient is not easily measurable. For speed regulation, a nonlinear robust speed controller is used in this work. The controller produces an estimate of the aerodynamic torque of wind turbine. The quality of this estimate improves with time. From that, a good estimate of power coefficient can be obtained. Simulations in MATLAB are executed with a model of a wind turbine based on its dynamic equations. From simulations, it can be seen that the one measurement SPSA method works very well for the wind turbine. It changes the tip speed ratio and blade pitch simultaneously, and the power coefficient reaches its maximum value quickly in a reliable manner. The power capture optimization is then implemented in FAST, a turbine simulation model created by NREL which is used to test the 5MW NREL reference turbine. From the results, it is evident that the wind turbine reaches the maximum power coefficient rapidly.

Doctor of Philosophy / Department of Mechanical and Nuclear Engineering / Youqi Wang / Textile composites are made from textile fabric and resin. Depending on the weaving pattern, composite reinforcements can be characterized into two groups: uniform fabric and near-net shape fabric. Uniform fabric can be treated as an assembly of its smallest repeating pattern also called a unit cell; the latter is a single component with complex structure. Due to advantages of cost savings and inherent toughness, near-net shape fabric has gained great success in composite industries, for application such as turbine blades. Mechanical properties of textile composites are mainly determined by the geometry of the composite reinforcements. The study of a composite needs a computational tool to link fabric micro- and macro-geometry with the textile weaving process and composite manufacturing process. A textile fabric consists of a number of yarns or tows, and each yarn is a bundle of fibers. In this research, a fiber-level approach known as the digital element approach (DEA) is adopted to model the micro- and macro-geometry of fabric and fabric reinforced composites. This approach determines fabric geometry based on textile weaving mechanics. A solver with a dynamic explicit algorithm is employed in the DEA. In modeling a uniform fabric, the topology of the fabric unit cell is first established based on the weaving pattern, followed by yarn discretization. An explicit algorithm with a periodic boundary condition is then employed during the simulation. After its detailed geometry is obtained, the unit cell is then assembled to yield a fabric micro-geometry. Fabric micro-geometry can be expressed at both fiber- and yarn-levels. In modeling a near-net shape fabric component, all theories used in simulating the uniform fabric are kept except the periodic boundary condition. Since simulating the entire component at the fiber-level requires a large amount of time and memory, parallel program is used during the simulation. In modeling a net-shape composite, a dynamic molding process is simulated. The near-net shape fabric is modeled using the DEA. Mold surfaces are modeled by standard meshes. Long vertical elements that only take compressive forces are proposed. Finally, micro- and macro-geometry of a fabric reinforced net-shape composite component is obtained.

Master of Science / Department of Mechanical and Nuclear Engineering / Mohammad H. Hosni / Byron W. Jones / With the increasing number of passengers traveling on commercial aircraft, it is important to mitigate the possibility of diseases and contaminants spreading throughout aircraft cabins and becoming harmful to the health of passengers. The ventilation system on a Boeing 767 aircraft is designed to create lateral flow to isolate contaminants to a single row of the cabin and remove the harmful air quickly. There are many variables that can influence the airflow patterns inside the cabin. The thermal plumes created by occupants are one of the variables investigated in this experimentation. Another special case investigated is the transport of gases in the cabin when the ventilation air is eliminated. Experimentation is performed in a mock-up Boeing 767 cabin. The mock-up enclosure consists of 11 rows and 7 columns of seats in each row. Ventilation apparatus, seating, and cabin dimensions used for testing are all representative of an actual aircraft. Thermal manikins are placed in the cabin seats to simulate the heat load from a seated person. A mixture of carbon dioxide (CO²) and helium (He) is injected into the cabin as a tracer gas to simulate the release of contaminants. The CO² concentration is measured by analyzers placed at the cabin inlet, exhaust, and seat of interest. The tracer gas can be injected and sampled at any of the 77 seats. In order to determine the effects of passenger density, testing is performed with maximum occupant load and repeated with half of the passenger load. Tracer gas is injected in three locations of the cabin and sampled in 32 seats for each injection seat. The testing revealed a significant effect of passenger load on airflow patterns. To determine the effects of removing the ventilation air, the cabin is supplied with 1400 cfm of outdoor air at 60°F for three hours to bring the cabin to a steady state temperature. Then, the supply air is shut off, and tracer gas is injected into the cabin and the CO² concentration is sampled at 12 locations throughout the cabin. It was found that contaminants are still transported throughout the cabin without the ventilation air.

Master of Science / Department of Mechanical Engineering / Donald Fenton / The need to store and access electronic information is growing on a daily basis as more and more people conduct business and personal affairs through email and the internet. To meet these demands, high energy density data centers have sprung up across the United States and around world. To ensure that vital data centers run constantly, proper cooling must be maintained to prevent overheating and possible server damage from occurring. Emergency cooling systems for such systems typically utilize traditional batteries, backup generator, or a combination thereof. The electrical backup provides enough power to support cooling for essential components within the data centers. While this method has shown to be reliable and effective, there are several other methods that provide reliable emergency cooling at a fraction of the cost. This paper address the lack of information regarding the initial, operation, and maintenance costs of using Thermal Energy Storage (TES) tanks for emergency cooling. From research and various field examples, five emergency cooling system layouts were designed for various peak cooling loads. Looking at the different cooling loads, components, and system operations an economic evaluation of the system over a 20 year period was conducted. The economic analysis included the initial and maintenance costs of each system. In an effort to better understand power consumption of such systems and to help designer’s better estimate the long term costs of TES tanks systems, five layouts were simulated through a program called TRNSYS developed for thermal systems. To compare against current systems in place, a benefit to cost ratio was done to analyze TES versus a comparable UPS. The five simulated systems were one parallel pressurized tank, one parallel and one series atmospheric tank, one parallel low temperature chilled water, and one series ice storage tank. From the analysis, the ice storage and pressurized systems were the most cost effective for 1 MW peak cooling loads. For 5 MW peak cooling loads the ice storage and chilled water systems were the most cost effective. For 15 MW peak loads the chilled water atmospheric TES tanks were the most cost effective. From the simulations we concluded that the pressurized and atmospheric systems consumed the least amount of power over a 24 hour period during a discharge and recharge cycle of the TES tank. From the TRNSYS simulations, the ice storage system consumed 22 – 25% more energy than a comparable chilled water system, while the low temperature storage system consumed 6 – 8% more energy than the chilled water system. From the benefit-cost-ratio analysis, it was observed that all systems were more cost effective than a traditional battery UPS system of comparable size. For the smaller systems at 1 MW the benefit-cost-ratio ranged between 0.25 to 0.55, while for larger systems (15 MW) the ratio was between 1.0 to 3.5 making TES tanks a feasible option for providing emergency cooling for large and small systems.

Doctor of Philosophy / Department of Industrial & Manufacturing Systems Engineering / Zhijian Pei / Donghai Wang / Both the U.S. and world economies have been depending on petroleum based liquid transportation fuels (such as gasoline, diesel, and jet fuels), which are finite and nonrenewable energy sources. Increasing demands and concerns for the reliable supply of liquid transportation fuels make it important to find alternative sources to petroleum based fuels. One such alternative is cellulosic ethanol. Research, development, and production of cellulosic ethanol have received significant support from both the U.S. government and private investors. However, several technical barriers have hindered large-scale, cost-effective manufacturing of cellulosic ethanol. One such barrier is related to the low density of cellulosic feedstocks, causing high cost in their transportation and storage. Another barrier is the lack of efficient pretreatment procedures, making pretreatment one of the most expensive processing steps and causing efficiency in the subsequent enzymatic hydrolysis to be very low. There is a crucial need to develop more cost-effective processes to manufacture cellulosic ethanol. Ultrasonic vibration-assisted (UV-A) pelleting can increase not only the density of cellulosic feedstocks but also sugar and ethanol yields. It can help realize cost-effective manufacturing of cellulosic ethanol. This PhD research consists of eleven chapters. Firstly, an introduction of this research is given in Chapter 1. Secondly, a literature review on ultrasonic pretreatment for ethanol manufacturing is given in Chapter 2 to show what has been done in this field. Thirdly, a feasibility test on UV-A pelleting of cellulosic biomass is conducted in Chapter 3. Comparisons of the pellet density and sugar yield are also made between pelleting with and without ultrasonic vibration. Next, effects of process variables (such as biomass moisture content, biomass particle size, pelleting pressure, and ultrasonic power) on output variables (such as pellet density, durability, stability, and sugar yield) have been studies in Chapters 4~6. Chapter 7 compares sugar yields between two kinds of materials: pellets processed by UV-A pelleting and biomass not processed by UV-A pelleting under different combinations of three pretreatment variables (temperature, processing time, and solid content). Next, mechanisms through which UV-A pelleting increases sugar and ethanol yields are investigated in Chapters 8 and 9. Then, a predictive model of pellet density is developed for UV-A pelleting in Chapter 10. Finally, conclusions are given in Chapter 11.

Master of Science / Department of Mechanical and Nuclear Engineering / Zhongquan Zheng / This study investigates a method using computational fluid dynamics (CFD) to model low-elevation atmospheric conditions. There are three goals in this research: to analyze the wind behavior downwind from buildings and trees, to validate the accuracy of the simulations by comparing wind measurements to the simulation for a specific site, and to find a relationship between the wind speed and the power output of a small-scale wind turbine. The first goal is to define a proper CFD model for buildings and trees. The trends in the Strouhal number are found to correlate to changes in building height and the wind resistance of a tree as supported in literature, with minor differences with the addition of a tree. The second goal of this study is to model an actual low-elevation environment to compare the energy output predictions for a small-scale wind turbine versus traditional methods. The simulations are compared to on-site wind measurements at a suburban wind turbine, recorded by the rotor and two anemometers installed on the wind turbine tower. The measurements and simulations presented in this study show an improvement in the accuracy in the estimation of the energy output of a wind turbine versus using traditional methods involving high-elevation wind maps. The third goal is to provide a relationship between the wind speed and the power output of a small-scale wind turbine. To accomplish this task, system identification is implemented. The traditional auto-regressive model with exogenous input variables (ARX), its moving average counterpart (ARMAX), and the output error (OE) model are compared in this study. It is found that the transfer function provided by the ARX model most sufficiently estimates the power output of the studied wind turbine, with power output accuracies of 83%. With all three goals addressed, the feasibility of small-scale wind turbines in different low-elevation environments is assessed. In accomplishing these tasks, the siting of a small-scale wind turbine can be optimized qualitatively and quantitatively.

Master of Science / Department of Mechanical and Nuclear Engineering / Kirby S. Chapman / The increased emissions monitoring requirements of industrial gas turbines have created a demand for less expensive emissions monitoring systems. Typically, emissions monitoring is performed with a Continuous Emissions Monitoring System (CEMS), which monitors emissions by direct sampling of the exhaust gas. An alternative to a CEMS is a system which predicts emissions using easily measured operating parameters. This system is referred to as a Parametric Emissions Monitoring System (PEMS). A review of the literature indicates there is no globally applicable PEMS. Because of this, a PEMS that is applicable to a variety of gas turbine manufacturers and models is desired. The research presented herein includes a literature review of NOx reduction techniques, NOx production mechanisms, current PEMS research, and combustor modeling. Based on this preliminary research, a combustor model based on first-engineering principles was developed to describe the NOx formation process and relate NOx emissions to combustion turbine operating parameters. A review of available literature indicates that lean-premixed combustion is the most widely-used NOx reduction design strategy, so the model is based on this type of combustion system. A review of the NOx formation processes revealed four well-recognized NOx formation mechanisms: the Zeldovich, prompt, nitrous oxide, and fuel-bound nitrogen mechanisms. In lean-premixed combustion, the Zeldovich and nitrous oxide mechanisms dominate the NOx formation. This research focuses on combustion modeling including the Zeldovich mechanism for NOx formation. The combustor model is based on the Siemens SGT-200 combustion turbine and consists of a series of well-stirred reactors. Results show that the calculated NOx is on the same order of magnitude, but less than the NOx measured in field tests. These results are expected because the NOx calculation was based only on the Zeldovich mechanism, and the literature shows that significant NOx is formed through the nitrous oxide mechanism. The model also shows appropriate trends of NOx with respect to various operating parameters including equivalence ratio, ambient temperature, humidity, and atmospheric pressure. Model refinements are suggested with the ultimate goal being integration of the model into a PEMS.