System identification techniques were applied to obtain mathematical expressions relating column flotation operating parameters. The empirical data were derived by making individual step changes in tailings rate, wash water rate, feed rate, frother addition and air rates, while measuring the pulp level and air holdup responses. The dynamic behavior of the flotation column is also analyzed for the conditions at which the tests were performed. After converting these mathematical equations into continuous and discrete transfer function matrices, several control algorithms were simulated on this column flotation empirical representation. Suggestions are provided to deal with the complexities of the process when designing a control scheme. / Master of Science

The crystallization behavior of poly(p-phenylene sulfide) PPS has been studied. Two PPS samples with < M_w > = 43K and < M_w > = 83K were fractionated to remove low molecular weight oligomers yielding fractionated PPS samples with < M_w > = 57K and < Mw > = 113K. The fractionated samples were then treated with an ion exchange process to allow control over the nature of endgroup counterion. Isothermal rates of hulk crystallization were analyzed as a function of n1olecular weight of PPS, fractionation, and chen1ical nature of the endgroup counter-ion. Additionally the spherulitic growth rates and nucleation densities were studied as a function of the chemical nature of the endgroup counter-ion for fractionated PPS with < M_w > = 57K. As a function of endgroup counter-ion, crystal growth rates and overall rates of crystallization decreased in the following order: H > Zn > Na. No significant trends in nucleation density as a function of endgroup counter-ion were observed. Bulk crystallization rates, nucleation density, general morphological features, equilibrium melting point, and abso1ute crystalline percentage of linear and starbranched nylon-6 have been studied as a function of branch-point functionality and crystallization temperature. Overall bulk crystallization rates were described in terms of the Avrami equation. The crystallization half-times of star-branched nylon-6 as a function of <i>supercooling</i> were reduced relative to those of linear nylon-6 of comparable molecular weight. Irregularities in lame1lar structure were implied by SAXS experiments on samples with branch-point functionality as low as three. The effects of changing the block length of poly(dimethylsiloxane), PSX, blocks in poly(etheretherketone)-poly(dimethylsiloxane), PEEK-PSX, multi block copolymers and their ketimine precursors, PEEKt-PSX, were investigated. A dran1atic dependence on thermal history prior to crystallization was observed in both block copolymer systems. Block copolyn1crs crystallized from the glassy state obtained a much higher degree of crystallinity at a n1uch faster rate of crystallization than copolymers crystallized from the melt at identical crystallization temperatures. Thermal and mechanical behavior of bisphenol-A polycarbonate was studied as a function of thermal history and absorbed mass fraction of C0₂. Mechanical testing and gas absorption experiments were performed on physically aged and unaged polycarbonate. Gas absorption studies indicated that although initial diffusion was somewhat retarded in the aged samples, both aged and unaged polycarbonate samples showed identical equilibrium absorbed gas values at 6500KPa and identical gas desorption behavior. Absorbed C0₂ was shown to plasticize polycarbonate. Polymeric liquid crystalline foams were produced from star-branched polyoxybenzoate-po1yoxyphenoxybenzoate copolymers (POB-co-POPB) via gas supersaturation followed by thermal blowing. Solid state C0₂ gas absorption well as melt processability of star-branched POB-co-POPB was increased relative to the linear polymer of comparable molecular weight. While C0₂ blown foams of linear POB-co-POPB produced by the gas supersaturation technique had a relatively high density and showed highly anisotropic bubble growth, well defined, nearly isotropic foams of star-branched POB-co-POPB with mean cell size from 200 to 400μ were made using the gas supersaturation technique. / Ph. D.

The concept of soft-switching pulse-width-modulated (PWM) technique was proposed aimed at combining the advantages of both the conventional PWM technique and the resonant technique. This work presents four new families of soft-switching PWM converters: the zero-voltage-switched (ZVS) PWM converters, the zero-current-switched (ZCS) PWM converters, the zerovoltage- transition (ZVT) PWM converters, and the zero-current-transition (ZCT) PWM converters. The family of ZVS- and ZCS-PWM converters are developed to improve the performance of the ZVS and ZCS quasi-resonant converters, respectively. The principles of operations of these two families of converters are presented, and the merits and limitations are assessed. A number of experimental converters are breadboarded to verify the theoretical analysis. Both the ZVT-PWM and ZCT-PWM techniques use the concept of shunt resonant network to achieve soft-switching. In this way, the new converters achieve soft-switching without increasing the voltage and current stresses of the power switches and diodes. By using the boost topology as an example, a complete dc analysis of the ZVT-PWM and ZCT-PWM converters is presented, and the dc Voltage-conversion ratio characteristics are derived. Design trade-offs are examined, and design procedures are established. The theoretical analysis and novel features of the proposed converters are verified on a number of breadboarded converters. Finally, the typical small-signal characteristics of the ZVT -PWM converters are analyzed and verified experimentally by using the boost converter as an example. / Ph. D.

An experimental study to determine the effects of riblets on the performance of the supersonic through-flow fan (STF) cascade blades was performed. The two-dimensional cascade was tested in the Virginia Polytechnic Institute and State University intermittent wind tunnel facility, where the Mach and Reynolds (based on chord) numbers were 2.36 and 4.8 x 10⁶ , respectively. Three different V-grooved riblet heights were tested on the blades: 0.023, 0.033, and 0.051 mm. Riblet testing were conducted at design incidence as well as at off-design conditions (incidence angles: +5, -10 deg). The riblet effect on the performance of the STF blades was determined by measuring the total pressure profile downstream of the cascade and integrating this total pressure to obtain an overall mass-averaged loss coefficient. The riblet loss coefficient was compared with the loss coefficient of a control test case where an equivalent thickness of smooth material is applied to the blades. Results show that, at the design incidence, the 0.033 mm height riblets provided the optimal benefit, with a reduction of 8.5% in the loss coefficient compared to the control case. Smaller effects were measured at the off-design conditions. Shadowgraph pictures were taken to study the effect of riblets on the turbulent transition location on the blades surfaces. At design incidence, the shadowgraphs revealed that the optimum height rib lets delayed the transition location on the suction surface of the blades. Therefore, it was concluded that for the 0.033 mm height riblets the decrease in the cascade's loss coefficient was the result of delayed transition in addition to a decrease in turbulent viscous losses. A numerical simulation was conducted to investigate both rib let effects on the STF blades. The numerical study showed that only the combination of the two riblet effects was able to produce a decrease in loss coefficient that was observed experimentally. Results from the numerical study indicate, that at design incidence, 2/3 of the rib let benefit is attributed to the delayed transition effect on the blades with the other 1/3 resulting from a decrease in turbulent viscous losses. / Ph. D.

This case study determined whether there was a difference in perceptions among principals from different quality schools about the influence desegregation, supplemental financial assistance, magnet thematic activities, and increased educational opportunities had on quality in education. Since there was no significant difference in perceptions, this study also examined how the factors influenced the principals' beliefs about educational quality. Two hundred fifty-five elementary and secondary principals from four different types of organizational structures in ten states participated in the study. The participating principals responded to the Survey of principals' Attitudes Regarding Education (SPARE) which constituted the data for this investigation. The degree of achievement, student dropout, violence, and racial balance were used to determine school quality and constituted the dependent variables. Principals' perceptions about desegregation, thematic activities, supplemental financial assistance, and increased educational opportunities constituted the independent variables. The findings of this study revealed the following conclusions: (1) There was no significant difference among principals from lower, average, and higher quality schools regarding their perceptions about the influence finance, desegregation, thematic activities, and educational opportunities had on the quality of education in a school. (2) Principals believed three of the four factors including desegregation, finance, and increased educational opportunities were necessary for a school to offer quality education. (3) Principals perceived student achievement could be increased without additional financial resources; however, educational quality could not be attained without additional funds for desegregation-related services and for programs designed to improve educational opportunities for students. Based upon information obtained from the survey instruments, principals believed a quality instructional program required a curriculum with special activities designed to improve student knowledge while increasing educational opportunities for a racially diverse student body. They believed schools should be given more funds for more specialized activities. Although magnet school thematic activities accomplished the same purpose, their costs were too high for most boards of education to finance. Regular school activities would be as effective as magnet school thematic activities with similar supplementary financial resources. / Ed. D.

The quest for developing high performance materials, originally responsible for the birth and growth of the composites industry, is now responsible for the drive to produce composites which outperform those used today. A great deal of this interest lies in the use of thermoplastic polymers as the composite matrix. Thermoplastics offer the potential of improved mechanical performance, greater solvent resistance, better impact tolerance, as well as cost saving processing attributes such as infinite shelf life, the ability to be recycled, short processing time and repairability. Unfortunately, these advantages come at the expense of a very high melt viscosity which makes the prepregging process, where the fiber and polymer are combined, very difficult. To overcome this and associated difficulties a detailed analysis of thermoplastic matrix composite processing is developed. The primary area of focus involves the development of an electrostatic dry polymer powder prepregging facility. This unit is capable of the production of high quality towpreg from thermoplastic polymers and reinforcing fibers. Two approaches, statistical and stochastic, were employed to model the process of polymer deposition on the fibers in the prepregging process. These models are used to optimize the production of towpreg. A modification of the prepregging facility allows for the production of towpreg from very small (15 g) samples of polymer. This is extremely useful for analysis and performance verification of state of the art polymer systems. / Ph. D.

A software system is an aggregate of communicating modules, and there are several different types of communication among these modules (direct, indirect, and global). Therefore, understanding the interfaces among these modules can characterize the system and are a major factor in the system's complexity. These interfaces could possibly also show and predict inadequacies in the reliability and maintenance of a system. Interfaces are defined early in the development life cycle at a detailed or high level design stage. Knowing that these interfaces exist and their structure leads us to measure them for an indication of the designed interface complexity. This designed interface complexity can then be utilized for software quality assurance by allowing users to choose from among several designs. With data provided by an Ada software developer, the interface complexity metrics correlated with established metrics, but also found complex interfaces that established metrics missed. / Ph. D.

Experiments were conducted to explore physiological, pathological, and immunological aspects of stress in chickens. Mechanisms of habituation and their association with adrenal steroids were also studied. Degree of stress experienced during perinatal and neonatal stages was evaluated. In both stages, heterophilia and lymphopenia were noted and post-hatching heterophil to lymphocyte (H/L) ratios declined in a quadratic manner with age. Feed restriction elicited a similar stress response, as indicated by H/L ratios in dwarfs and normals with the effect of adapting to fasting having dissipated between 12 to 16 days. Resistance to Eimeria tenella infection was greater in normals than dwarfs, and for fasted than ad libitum fed chicks. Neither genotype nor feeding regimen had an effect on antibody response to SRBC. Frustration induced by thwarting of feeding resulted in a dramatic elevation in HIL ratios and conspicuous behavioral responses, suggesting responses to fasting may be physiologically and psychologically modulated. Provision of sand during feed withdrawal attenuated physiological responses, possibly via alteration of psychogenic components. An experiment was designed to evaluate the impact of mild feed restriction at several ages on heat tolerance. Normal chicks feed-deprived during the neonatal stage had lower H/L ratios, improved resistance to marble spleen disease and growth during heat treatment than those fed ad libitum or fasted at older ages. A companion experiment examined habituation and concomitantly the role of glucocorticoids in stress response. Neonatal fasting-induced stress responses without concurrent increases in adrenal steroidogenesis did not enhance ability of normal chicks to withstand high ambient temperatures. The hypothesis regarding modulatory roles of corticosteroids in neonatally-elicited adaptation was affirmed in a subsequent study. Feed withdrawal for 8 or 24 h at 36 days of age did not induce either H/L or plasma corticosterone response of chicks stressed early in life. In contrast, neither criterion provided evidence for adaptation to feed deprivation by neonates treated with an adrenal steroidogenesis blocker during the fast. corticosterone mediated inhibitory feedback regulation of the hypothalamic - pituitary - adrenal axis which down-regulates ACTH response was also evident. / Ph. D.

Seed priming is a controlled hydration process, followed by dehydration, that allows pregerminative metabolic activity to proceed without germination. The objective of this research was to investigate the effects of priming on intrinsic characteristics of seed germination including temperature, water, and development, in order to understand how priming affects the germination of broccoli (<i>Brassica oleracea var. italica</i> Plenck) seeds. Priming of broccoli seeds consistently improved germination and emergence rate in the laboratory and field and was related to the accumulation of a specific level of hydropriming units expressed in MPa*h. Priming reduced the sensitivity of seed germination to temperature and increased the temperature range of germination but did not lower the minimum temperature for germination. Primed seeds leaked less electrolytes at supraoptimal temperatures (≥ 35°C) compared to nonprimed seeds. In the field, primed seeds produced a greater plant stand and yield under stressful emergence conditions. Under optimal conditions in the field for stand establishment, the advancement in emergence of primed seeds did not carry over to earlier, greater yields. Matric priming, using calcium silicate as the carrier in the ratio 1.0:0.8:1.8 (seed:carrier:water; by weight) for 7 d at 20°C, was superior to osmotic priming using polyethylene glycol (pEG 8000) at -1.2 :MPa in nearly all variables examined. This may be attributed to reduced respiration during priming of seeds in PEG or nutrient uptake by seeds in calcium silicate. The testa was observed to be a barrier to broccoli seed germination. Priming-induced changes to the physical characteristics of broccoli seeds included increased volume (32%) and an irreversibly expanded, and weakened testa with some minute cracking near the area where the radicle emerges. Primed seeds germinated faster, in part by maintaining a lower hydrotime constant, and thus exhibited a greater progression towards germination per unit water potential at a constant temperature compared with nonprimed seeds. It was hypothesized that, since the testa threshold was reduced after priming by expansion and formation of free spaces, the reversibly expanded embryo of primed seeds does not become immediately appressed to the testa upon rehydration. Thus the yield threshold component controlling the rate of germination of primed seeds is lower by the amount of the testa threshold. The priming effect is more than just reducing the yield threshold as indicated by a significant invigoration of seeds with split testae. Despite the increased volume as a result of the formation of free spaces, primed broccoli seeds did not imbibe more water or have a greater turgor at full hydration. Priming did not lower the minimum water potential allowing germination, and primed seeds did not plateau in water uptake but, instead, moved immediately from imbibition to expansive growth. Priming improved the germination rate of broccoli seeds at all stages of maturity with the most significant effects at stages before attainment of maximum dry weight. Dry storage of broccoli seeds at harvest able maturity (≥ 56 days after pollination) did not improve germination, indicating a lack of post harvest dormancy. / Ph. D.

This research shows that a model that concurrently determines product, production, and capacity can provide better results than a model that makes these decisions sequentially. Three versions of a life-cycle complete concurrent model and corresponding sequential model that solve product, production, and capacity planning problems are mathematically formulated and compared through an hypothetical example. All versions of the concurrent and sequential models consider three types of work centers and corresponding types of capital budgeting projects. Each version of the concurrent and sequential models considers a specific type of capital budgeting project. These are: (1) projects that may be partially accepted with no upper limit, (2) projects that may be partially accepted with an upper limit of one, and (3) projects restricted to values of zero and one. An hypothetical example spanning twelve fiscal periods that included five products, two product sub-groups, one pre-production/post-support project, three production projects, and one support project was developed to enable comparison of the models. Nine comparisons between each of the three versions of the concurrent and sequential models were conducted resulting in a total of 27 comparisons. Comparisons of the models required the development of five computer programs. Four of these programs provided Mathematical Programming System (MPS) formatted mixed integer linear programs (MIP) that were solved using the LINDO/386 optimization program. The fifth program read output from LINDO/386 and provided the future worth of the sequential model. Results obtained from tests of the concurrent and sequential models were compared based upon future worth, capital budgeting projects funded, products funded, production and inventory quantities, and regular and overtime labor levels. For every combination of product sub-group demand and capital budgeting constraint type, the concurrent model always provided the highest future worth. In addition, while both models funded the same set of products, the concurrent model provided more level labor, production and inventory quantities, and support for products. Following these comparisons, it was concluded that, for any capital budgeting constraint type and product sub-group demand, a concurrent model can provide better planning results than a sequential model. / Ph. D.