The iron oxides are a group of materials with geological, biological, and technological importance. A thermodynamic understanding of these materials is important because it provides information about their relative stabilities, chemical reactivity, and transformations. This study provides the heat capacity of a nanocrystalline magnetite (Fe3O4) sample, bulk hematite (α-Fe2O3), nanocrystalline hematite, akaganéite (β-FeOOH), and lepidocrocite (γ-FeOOH) at temperatures as low as 0.5 K. These measurements were fit to theoretical functions at temperatures lower than 15 K, and the respective thermophysical properties of these materials are discussed. Also the molar entropies of bulk hematite and hydrous nanocrystalline hematite as well as hydrous akaganéite are given. Finally, a ferritin protein powder was prepared for heat capacity measurements by reconstituting the iron core in the presence of an imidazole buffer. This method allowed the introduction of almost 3000 iron atoms into each protein. Heat capacity measurements of apoferritin and the reconstituted ferritin sample are anticipated in the near future with plans to compare the heat capacity of the mineral core to that of other nanocrystalline iron oxides and oxyhydroxides.

Niger, a northwest African country with several systemic barriers to development, has made education a priority. In an effort to improve the national education system, Niger has implemented a decentralization program. This study examines the perceptions of local school actors concerning this decentralization policy, which prescribes improving access and quality to education and strengthening institutional capacity. Local interviews and an analysis of relevant policy documents reveal limited policy implementation at the local level accompanied by a lack of state capacity, accountability and responsiveness to local school needs. Moreover, interviewees perceive a decline in education quality due to these reforms. While policy review documents focus on building institutional capacity at the central and regional government levels, the locality examined has responded as best it can to the needs of its schools. These local efforts are hampered by few resources, limited capacity and understanding of the importance of education by citizens, as well as a mistrust in government institutions, including schools, among local community members. Community leadership, development of participatory public space and trust building, could improve local education capacity to a certain extent, however, strong central government that provides additional resources and builds the capacities of school staff is necessary. / Master of Urban and Regional Planning

The WMC-gF relationship has been attributed to attentional control by some, and to a learning-based account by others. The current study explores inconsistencies in solving structurally-identical problems and how such factors may explain the WMC-gF relationship. Participants completed multiple versions of the same visual-analogies problems, with some problems sharing surface features and others looking vastly different, to test the ability to generalize a rule. In addition, subsequent iterations were shown either immediately after the first presentation, after two intervening items (second presentation), or after at least 10 intervening items (third presentation). Performance on second-presentation items supported both attention and learning-based accounts and performance on third-presentation items supported only a learning-based account. Furthermore, surface similarities interacted with third-presentation item accuracy and WMC, with a stronger relationship for dissimilar looking items. These findings suggest that the ability to learn and generalize rules throughout a task may largely contribute to the WMC-gF relationship.

No

<p> Wireless optical communication systems have the potential of establishing secure high data rate communication links. In order to realize the ultimate promise of these links, channel modeling and communication algorithms must be developed. This thesis addresses free-space optical (FSO) system design and provides novel contributions in four major areas: 1) channel modeling, 2) channel capacity and optimal signal design, 3) signaling algorithms, and 4) formal methods to jointly design code rate and beamwidth for FSO systems.</p> <p> A novel statistical channel model taking into account atmospheric and misalignment fading is developed that generalizes the existing models and accounts for transmitter beamwidth. The channel capacity is analyzed under average and peak optical power constraints and a new class of non-uniform discrete input distributions are developed with mutual information that closely approaches the channel capacity. Algorithms to realize the proposed non-uniform signaling and achieve the promising rates are also presented. Numerical simulations are conducted with finite length low density parity check codes showing significant improvement in system performance. Finally, the developed signaling is applied to FSO channels considering the above impairments. Beamwidth optimization is considered to maximize the channel capacity subject to outage. It is shown that a rate gain of 80% can be achieved with beamwidth optimization.</p> / Thesis / Doctor of Philosophy (PhD)

This thesis investigates the interactions of several characteristics of serial production lines that contribute to production line performance. A full factorial experimental design of computerized simulations is conducted with three levels of downtime, four levels of variability, three levels of line length, three levels of constraint location, three levels of work-in-process, and six levels of protective capacity. This study enlarges upon recent four-workstation investigations and extends the knowledge to longer production lines. Some generalizations for the amount and location of protective capacity are drawn from the results, as a guide for process improvements and new production line design. An approximating regression model is constructed for prediction of cycle time outcomes with various values of contributing factors.

No description available.

No description available.

No description available.

No description available.