Industrial manufacturing processes can experience a variety of changes to important quality characteristics as a result of tool breakage, tool wear, introduction of new raw materials, and other factors. Statistical process control charts are often used to monitor for changes in quality characteristics for manufacturing processes. The control chart computes a statistic based on measured observations of the process and compares it to control limits. When the statistic exceeds a control limit, the control chart signals that the process is out-of-control. Quality engineers would then search for the special cause responsible for the change in the process. Rapid detection by the control chart is important to minimize the production of poor quality items as a result of an out-of-control process. Control charts that detect changes rapidly can therefore save critical process down-time and expense. This research investigates several new control charts related to the commonly used cumulative sum (CUSUM) control chart. The new control charts use control limits that change as a function of the number of process observations instead of remaining constant. The performances of these new control charts are compared to several other charts including the Shewhart chart, the CUSUM, and the exponentially weighted moving average charts. Compared to the standard CUSUM, the proposed control charts can detect a change in a process more rapidly for a given range of shifts in the mean of a process. The proposed control charts offer a flexibility to quality engineers for better optimization of monitoring schemes for manufacturing processes. / A Thesis submitted to the Department of Industrial Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2004. / March 31, 2004. / Average Run Length, ARL, Generalized Sequential Probability Ratio Test, Quality Engineering, Control Charts, CUSUM / Includes bibliographical references. / Joseph J. Pignatiello, Jr., Professor Directing Thesis; Okenwa I. Okoli, Committee Member; James R. Simpson, Committee Member.

The Florida Department of Transportation has initiated and funded development of electronic crash and citation reporting in Florida using the TraCS (Traffic and Criminal Software) platform. The TraCS system is application software that is a customizable data collection system that can be used by law enforcement and motor vehicle agencies to collect crash data. TraCS is combined with laptop computers, one or more personal computers (PC) in a central office. Peripherals such as image/bar code scanners and mobile printers, and data communications are in conjunction with TraCS to provide officers with all of the functionality needed to record and retrieve incident information wherever and whenever an incident occurs. The thesis objectives were to perform time study analysis and to investigate costs of current and proposed (TraCS-based) methods. The TraCS software was developed, and in the summer of 2003; seven Florida law enforcement agencies were selected to pilot and test electronic crash and citation reporting. The agencies were provided with required equipment and training to use TraCS software. The ride-alongs were performed with Jacksonville Sheriff's Office and Leon County Sheriff's Office (two of the pilot agencies). During the ride-alongs, the time taken to complete forms both with and without TraCS software was measured and the data analysis was performed. The study shows that the efficiency and accuracy of Florida traffic records was improved by using the electronic data collection system (TraCS). Data analysis showed that it takes less time to fill a crash report using TraCS compared to filling out a crash report manually on a paper form. On an average for the two vehicle crashes the time saved by using TraCS software to fill long form, short form, and driver exchange form were 11.7%, 11.3% and 8.3%, respectively. The time to fill a citation form using TraCS software was reduced by 13.6% from the time without TraCS. The software had best application when used in conjunction with the magnetic stripe reader for Florida driver license. The efficiency of officer's using TraCS differs based on the learning curve, equipment provided, and mindset of an officer. After suitable training the time taken to complete a report should decrease even further. / A Thesis submitted to the Department of Industrial and Manufacturing Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester, 2004. / November 3, 2004. / Tracs, Traffic Data, Data Collection System, Automation, Computerization, Law Enforcement Automation / Includes bibliographical references. / Lisa K. Spainhour, Professor Co-Directing Thesis; Yaw A. Owusu, Professor Co-Directing Thesis; Okenwa I. Okoli, Committee Member; Joseph J. Pignatiello, Jr., Committee Member.

The resin transfer molding (RTM) process has been used in the composite industry for decades. However, several issues still exist and impede its wide applications. Some design tools for RTM parts have been developed but a more efficient design environment is lacking. Race-tracking is a common phenomenon that makes prediction in actual production difficult and makes current deterministic optimal tooling design unrepeatable. This thesis integrates flow simulation and cost analysis modules together with database management system (DBMS) providing a prototype of the integrated design environment for RTM processes. Preform permeability, especially race-tracking permeability that significantly affects not only simulated but also experimental results, was the factor being investigated. This thesis introduces a statistical approach utilizing statistically distributed variables to explain the race-tracking permeability values. One-dimensional flow experiments were conducted to obtain the permeability values. Three types of distribution (gamma distribution, Weibull distribution and lognormal distribution) were chosen as candidates. Experimental data were fitted for the three distributions. A goodness-of-fit test was performed to find the one that best describes the experimental data. Taking into account the fact that the severe levels of race-tracking can be represented by statistically distributed variables, this thesis proposes an optimization approach to minimize the sensitivity of the mold design to uncertainty of race-tracking permeabilities by choosing the appropriate locations of gates and vents (robust tooling design). A sensitivity that indicates the process robustness was defined as objective and evaluated by RTMSim software both for 2D and 2.5D geometry. With the conclusion that the ratios of race-tracking permeability over average values can be described by Weibull distributed variables, a random number generator was employed to generate the input race-tracking permeability data for obtaining values of the objective. Locations of vents were determined via the assumption that vents should be assigned at the locations where flow ends to avoid dry spot formation. Locations of gate were optimized from most possible locations. / A Thesis submitted to the Department of Industrial and Manufacturing Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester, 2003. / October 23, 2003. / Robust Design, Flow Simulation / Includes bibliographical references. / Chuck Zhang, Professor Directing Thesis; Okenwa Okoli, Committee Member; Zhiyong Liang, Committee Member.

As composite materials gain wider acceptance within the transportation industry, it is pertinent to investigate the available coating processes with a view to reduce emissions and associated costs. Current painting processes are not only laborious and time consuming, but expensive and present safety issues. In-mold coating presents an avenue for eliminating on reducing theses problems. This work aims to determine the feasibility and develop a methodology for In-Mold Coating (IMC) components manufactured using the Resin Infusion between Double Flexible Tooling (RIDFT) process. Comparisons of the cost and efficiency between a component manufactured by RIDFT and then painted, and a component manufactured by RIDFT IMC process is given. It indicates a 46% savings in capital investment and 55% savings in time while using RIDFT IMC. The viability of in-mold coating RIDFTed components was investigated. This work-in-process, reports on successes and challenges presented, during the co-infusion of a polyurethane enamel paint (DuPont Imron 5000) and a vinyl ester resin (Derakane 470-45). In this thesis, RIDFT IMC is used to manufacture a double-layered structure consisting of a vinyl ester layer for structural function and a polyurethane enamel coating layer. The two liquids are either simultaneously or sequentially infused into a double flexible mold and are cured. Liquid separation is maintained by a separation layer. Twenty-four case studies have been made to determine the viability of the IMC process. Eight different separation layers were tested to find out the most promising separating material. The specification of the separation layer was made. Several test evaluations were conducted. A Differential Scanning Calorimeter (DSC) is used to test the cure cycle for the resin and the paint. Mechanical testing, Dynamic mechanical analysis (DMA) and tensile tests are used to evaluate the performance of the produced parts. An Environmental Scanning Electron Microscope (ESEM) was used to evaluate the microstructure of the components. In order to improve the flow of the fluids, several Flow Distribution Channels (FDCs) were made and tested to determine the best configuration. The experimental results demonstrate that in-mold coating by means of co-infusion may be a viable option for painting RIDFT components with determination of an appropriate separation layer. The paint used in this work dissolves the separation layer. Further work has been suggested to develop a paint formulation that will not dissolve the separation layer. / A Thesis submitted to the Department of Industrial Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Summer Semester, 2004. / June 7, 2004. / Co-Infusion, Paint, In-Mold Coating, Composites, Painting, RIDFT, LCM / Includes bibliographical references. / Okenwa Okoli, Professor Directing Thesis; Ben Wang, Committee Member; Zhiyong Liang, Committee Member; Chuck Zhang, Committee Member.

Polycarbonate (PC) is a transparent, impact resistant polymer that provides protection against breakage or intrusion. The mechanical toughness of PC is reported to be associated with the molecular motion of main chain molecules. The molecular motion is present upon exposure to impact and can therefore provide efficient dissipation of impact energy. PC has found wide usages in military and commercial applications. This thesis investigates processing and characterization of transparent PC/carbon nanotube (CNT) composite films. The reinforcing capability and efficiency of CNTs in nanocomposites have been studied intensely world-wide due to their exceptional mechanical properties. Not only do they provide stiffness and strength, but they also have been reported to impart fracture toughness when dispersed in polymer matrices. As cracks develop in nanocomposites, CNTs serve as the bridging nanofibrils that effectively retard crack propagation. In some cases, CNTs act as obstacles that obstruct the crack propagation paths, thus increasing the energy needed to be dissipated for further crack opening. This research deals with two important issues in multi-walled carbon nanotube (MWNT)-based composites that are seemingly trade-offs – dispersion of high-loading MWNTs and maintenance of optical transparency. Higher loading of MWNTs is desired for increased reinforcing effects; however, it is limited by the difficulties in achieving uniform dispersion in the polymer resin due to increased viscosity and the tendency of MWNTs to aggregate. At the same time, only a small addition of MWNTs (less than 1 wt.%) makes the resin turn dark, which defeats the advantage of the transparency that should otherwise be retained by PC. PC/MWNT composite films (up to 0.15 wt.% MWNT) were fabricated using a solvent-based film casting method. Several optimization schemes were adopted to determine the most suitable solvent and the concentrations of PC/MWNT in the solvent. Solvent-induced PC crystallization, which was evidenced by milky tinting in the produced films, was minimized by identifying and optimizing the process parameters, namely, PC/MWNT/solvent solution viscosity, casting temperature, and film thickness. Design of experiments technique was used to determine the combination of optimal parameters. / A Thesis submitted to the Department of Industrial Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Spring Semester, 2007. / April 12, 2007. / Multi-Wall Carbon Nanotube, Transparent Film, Polycarbonate / Includes bibliographical references. / Young-Bin Park, Professor Directing Thesis; Ben Wang, Committee Member; Zhiyong Liang, Committee Member; Okenwa Okoli, Committee Member.

Carbon nanotubes (CNTs) possess great potential for developing high-performance and multifunctional nanocomposites for a wide variety of applications. As the cost of producing CNT buckypaper, a thin film of CNT networks, continues to decrease while the quality increases, more users and companies are becoming interested in buckypaper for potential applications. Many of these applications, such as electromagnetic interference (EMI) shielding and fire retardant surface skins for fiber-reinforced composites or plastics, may not require buckypaper-based composites to be much stronger compared to fiber-reinforced composites. This means that there is a market for buckypaper even without its theoretical super strength, but desired functionality. There is however a number of challenges with the potential scale-up production of composite parts with affordable buckypaper materials usually made of low cost CNTs, such as multi-walled carbon nanotubes (MWNTs). Such buckypaper is usually very lightweight (10-25 g/m2), thin (10-20 microns), and fragile, hence even small variations and damage in the wet lay-up process can result in large quality variations in the final buckypaper composite. These variations include buckypaper volume fraction, resin rich areas, and contact between BP and other reinforcement materials etc. Thus, keeping the consistency of the resulting microstructure and quality of buckypaper composites is a very challenging issue. The objective of this project is to study nanostructure-permeability relationships of different types of buckypaper materials, and explore effective prepreg processes to make buckypaper composites with greatly increased consistency, quality, CNT weight fraction and uniformity in the resulting products. The experimental results show that buckypapers have very low permeability, about 8-12 orders lower than those of carbon fiber preform cases, and also sensitive to liquid polarity due to their nanoscale porosity and large surface area. Both solution and resin film transfer prepregging processes were studied to pre-impregnate buckypaper to achieve 50 wt. % CNT concentration. The late one showed better quality in the resultant nanocomposites, but difficult for high viscosity resins. Three case studies were also conducted to demonstrate quality and property consistency of buckypaper composites. / A Thesis submitted to the Department of Industrial and Manufacturing Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Summer Semester, 2010. / July 12, 2010. / Prepreg, Nanotube, Buckypaper / Includes bibliographical references. / Richard Liang, Professor Directing Thesis; Ben Wang, Committee Member; Okenwa Okoli, Committee Member.

The main focus of this study is to evaluate whether load following operation can be performed using the existing South African nuclear power plant. Of which, knowing that will assist in determining whether the addition of load following capabilities on the planned nuclear fleet is justifiable or not. In this report the relevance of Koeberg Nuclear Power Plant to adapt to the demand is examined and the effects on plant operation simulated. The report analyses the operation of the existing nuclear power plant (Koeberg units) in South Africa and describes the regulations that govern safe operation of the plant. The Koeberg plant is analyzed based on the current design i.e. operating as a baseload station. This allows a prediction of the Koeberg plant response to big load variations. The simulation results of the load variation are analyzed and the results used to make the conclusion that the Koeberg units are not capable of load following safely. Modifying the Koeberg units from being base load station to load following will require changing the Safety Analysis Report and therefore affect the Koeberg license NIL-01.

Abdominal aortic aneurysms affect 0.2% of the population and are closely associated with intraluminal thromboses (ILTs) that develop in the sac. Advanced imaging and treatment techniques are available, however there is room for improvement in the methods used to predict the outcome or necessity of surgical intervention. For a computational model to be useful in this clinical setting, it would need to incorporate relevant patient-specific data and prioritise simplicity and speed over exhaustive detail. This paper presents the details of such a model, for abdominal aortic aneurysms, particularly in the simplification of the coagulation biochemistry. Explicit modelling of the coagulation cascade is replaced with a patient-specific thrombin generation curve. This curve is defined by three values obtained from a blood test. Another key feature is the thrombosis growth model, which incorporates conversion of fibrinogen to fibrin, variation between clot core and shell, and mechanical lysis. The model generates ILTs with morphologies visually similar to those typically found in the body, however more work is required to refine and validate the mode.

Textural variability is a key component in addressing process challenges resulting from variability in the ore being mined. Textural variability arises from differences in the types of mineral grains present, their relative abundance and the type of interactions they have with one another. Increased textural variability is the largest contributor to mineral processing challenges in terms of mill throughput and flotation concentrate grades. Processing of ores with high textural variability often results in reduced throughput and the recovery of lower grade product, if low-grade material is not eliminated prior to arriving at the concentrator. Geometallurgy provides a powerful tool to manage ore variability better by using geological and metallurgical information during plant design and operation. The geometallurgical approach contributes towards minimising and controlling operational and technical risk of ore variability. Ore breakage characterisation is a pivotal part of geometallurgy which aims to quantify the relationship between the energy supplied for breakage and the size of the resultant progeny. The Julius Kruttschnitt Rotary Breakage Tester® (JKRBT) is an ore breakage characterisation device designed as a geometallurgical tool which can use both crushed ore and drill core samples. Drill core is especially important as it the material used for geometallurgical testing during exploration and resource definition. The JKRBT is more accurate and the test work is less time consuming than its predecessors. However, sample availability is a major concern when performing metallurgical testing as numerous tests need to be performed to get a complete view of the metallurgical response of the given ore type. This means that very little of the sample is available for ore breakage characterisation. The aim of this work is to understand the relationship between mineral texture and the ore breakage characteristics of several samples for both drill core and crushed ore which are the two sample types used at exploration and operational levels. The work is aimed at addressing the problem of sample availability in geometallurgical testing by developing an ore breakage characterisation test protocol that uses minimal sample to extract relative hardness indices. To assess the influence of mineral texture on the ore breakage characteristics, five different ore types were used. The ore was prepared by coring different size drill core and crushing using a jaw crusher. The samples were subjected to controlled single impact breakage tests using the JKRBT. A standard test consisted of 3 energy levels (low, medium and high) tested on 3-4 particle size fractions (small, medium, large and very large). The least particle protocol was developed using an ore which was considered to be fine-grained and homogenous. In developing the protocol, all the steps of the standard procedure were followed except the number of particles per test was progressively reduced from 30 particles to a threshold of 5 particles. From ore breakage characterization tests performed, it was observed that ore S was the most resistant to breakage while ore P was the least resistant to breakage. Ore A was found to be more competent than ore B and ore C but less competent that ore P. The differences in the ore breakage characteristics were attributed to the grain size distribution of the dominant constituent minerals for each sample. The findings were attributed to an increase in the surface area available for contact due to the decrease in grain size which results in less stress per unit area and thus more resistance to breakage. The hardness of an ore can be considered to be a function of the mineral hardness and its relative abundance. Using the relative mineral abundance and Mohs hardness scale, it was concluded that the more abundant the harder minerals in an ore, the more resistant to breakage the ore is. Drill core particles consistently produced a coarser progeny than crushed particles at the same conditions. The observed trend was attributed to differences in the particle shape between the crushed particles (angular) and drill core particles (cylindrical). Angular particles have a larger surface area exposed for energy absorption and therefore break more easily than drill cores. Chapter 5 showed that the proposed abridged ore breakage characterisation test that uses the minimal number of particles to extract ore breakage indices can be applied for both homogenous and heterogenous ore types. The results also show that the proposed least particles protocol can be used for ore breakage characterisation tests using both crushed ore and drill core particles. The number of particles can be reduced to as little as 10 particles per test while still obtaining the same ore breakage indices as those obtained from the standard procedure. The abridged protocol will be especially useful in situations where the amount of ore available for metallurgical testing is limited.

The growing need to curb greenhouse gas emissions has made low-temperature proton exchange membrane fuel cells (PEMFCs) more attractive for automotive application. One of the major problems facing PEMFCs is the sluggish kinetics of the oxygen reduction reaction (ORR). To further enable wide-scale commercialisation of PEMFCs for automotive applications, major improvements in the ORR catalyst are therefore needed. An in depth understanding of the ORR mechanism over Pt surfaces can enable rational approaches in the search for more active ORR catalysts. The ORR occurs over multi-faceted Pt nanoparticles which predominantly expose Pt{111} and Pt{100} facets. Most studies have modelled the overall ORR activity over multi-faceted surface assuming that the Pt{111} and Pt{100} facets are kinetically isolated. Density functional theory (DFT) studies have shown that Pt(111) surfaces can efficiently facilitate OH* hydrogenation to H2O* but not the hydrogenation of O* to OH*, whereas Pt(100) surfaces can facilitate O* hydrogenation to OH* better than OH* hydrogenation to H2O*. If O* intermediates can readily diffuse from Pt{111} to Pt{100} facets and OH* from Pt{100} to Pt{111} facets, the ORR activity on Pt{111} and Pt{100} facets of multi-faceted surfaces may no longer be limited by O* and OH* hydrogenation steps, respectively. This study uses DFT and microkinetic models to investigate the nature of inter-facet cooperation and how it influences the ORR activity under dry conditions, i.e. catalyst surface exposed to a gas mixture of 33% O2 and 67% H2 at 1 bar. Under these conditions, it is assumed that the Langmuir-Hinshelwood kinetics are dominant. Using DFT, the adsorption, diffusion and reaction energetics of various reaction intermediates and reaction steps were calculated. The Pt{111} and Pt{100} facets were modelled as Pt(111)-p(3x3) and Pt(100)-p(3x3) slabs, respectively. The edge was modelled using a rhombic nanowire model with alternating Pt{111} and Pt{100} facets. Edge sites were found to adsorb oxygen containing species strongly. Consequently, the diffusion barriers of O* and OH* from edge sites towards terrace sites were much higher than the diffusion on the terraces and even higher than the activation barrier for reaction in the ORR. Replacing the edge Pt atoms with Au and Ag atoms weakens the adsorption of both O* and OH* on edge sites. Microkinetic analyses of ORR requires the inclusion of lateral interactions, since otherwise a full coverage of the surface with O* is predicted. Higher ORR rates are obtained on Pt(100) surfaces and --(vi)-- Pt{100} facets than on Pt(111) surfaces and Pt{111} facets. The ORR activity on Pt(111) and Pt(100) is limited by O* hydrogenation at T < 480 K and O2* dissociation at high temperatures. The ORR pathway varies greatly over these surfaces. On Pt(111), the ORR follows a peroxyl pathway at T < 500 K and a dissociative pathway at T > 700 K. On Pt(100) surface H2O* is formed via O* hydrogenation to OH* followed by 2OH* coupling to H2O* and O*. The ORR activity on multifaceted Pt surfaces was shown to be dependent on the ratio of edge sites to Pt{111} sites Modelling the inter-facet exchange of ORR intermediates based on data generated using Au and Ag modified nanowires could improve inter-facet cooperation. The most interesting case was Ag modified systems where inter-facet exchange of OH* occurs at temperatures as low as 360 K. On these systems, the ORR pathway on Pt{111} involves OH* diffusion from edge followed by OH* hydrogenation to H2O*. No O2 adsorbs on the Pt{111} facet. Edge modification has the ability to selectively enable inter-facet exchange of some reaction intermediates whilst inhibiting others. Therefore, it should be explored in rational catalyst design.