The purpose of this study was to explore the relationships between weight cycling and dietary restraint, disinhibition, binge eating, emotional eating and body mass index (BMI) while controlling for their simultaneous effects in women selected from a general population. Moreover, the interrelationships between the measures of dietary restraint, disinhibition, binge eating, and emotional eating were examined. The Disinhibition Subscale (DS) and the Cognitive Restraint Subscale (CRS) of the Three Factor Questionnaire (TFQ), the Emotional Eating Subscale (EE) of the Dutch Eating Behavior Questionnaire, and the Bulimic Inventory Test (BITE) were administered to 350 women in accordance with the Dillman protocol, and BMI was calculated. Significant positive correlations between weight cycling and the CRS (r = 0.23), DS (r= 0.38), BITE (r =0.39), EE (r = 0.36), and BMI (r= 0.28) were found. Together these variables accounted for 26% of the total variance in number of weight cycles. The EE and BMI had the strongest effect on weight cycling. The next strongest effects were the BITE and CRS; although, they were not statistically significant. The results suggest that obese women who diet repeatedly may be successful at losing weight but are not successful at long-term weight loss maintenance due to factors such as emotional and binge eating. / Master of Science

A distributed power system (DPS) is made up of several subsystems. For example, a two-stage distributed power system is made of a source subsystem consisting of line conditioners and a load subsystem consisting of loads. Motor drives used as a load subsystem form an important type of load in distributed power system applications. The input impedance of the load subsystem is an important factor in designing and analyzing the performance and stability of a distributed power system. In this thesis, a typical three-phase inverter-fed ac motor drive is modeled, analyzed and the input impedance characteristics are studied for the first time. Motor drives are found to have unique input impedance characteristics due to their electromechanical nature. The influence of these characteristics on the distributed power system are analyzed. The unique interaction problems that these characteristics lead to are studied. It is shown that a distributed power system designed without taking into account the unique input impedance characteristics of motors might suffer from performance degradation or might even become unstable. Design guidelines to avoid this situation in a distributed power system that uses motor drives as a load subsystem are developed and presented. / Master of Science

The purpose of this study was to estimate the extent of petrified wood theft at two popular petrified wood sites at Petrified Forest National Park and to describe the nature of the thieves in terms of their level of moral reasoning. A model of noncompliant behavior was developed that incorporated the theory of reasoned action, the theory of planned behavior, the theory of attitude accessibility, and the four-compartment model of morality. The model proved valuable in understanding petrified wood theft. Data were collected by non-obtrusive observations. brief on-site interviews, and an eleven-page mailback questionnaire. In total, 226 park visitors (125 thieves and 101 non-thieves) were observed. Of these, 145 (77 thieves and 68 non-thieves) agreed to be interviewed and were given a questionnaire. Altogether, 116 (61 thieves and 55 non-thieves) returned a completed questionnaire for a combined response rate of 80.9 percent. A total of 456 park visitors, who were not observed, were also contacted and given a mailback questionnaire. Of these, 365 returned a completed questionnaire for a 80.0 percent response rate. An average of 1.02 thefts were observed per hour of observation at the two petrified wood sites. And, 1.22 percent of visitors at the two petrified wood sites collected at least one piece of petrified wood. An average of 1.6 pieces of petrified wood were taken per theft. These findings emphasize the seriousness of petrified wood theft by park visitors. Thieves were found to be significantly more thoughtless than non-thieves as measured by an impulsiveness scale. Also, thieves scored significantly higher than non-thieves on the preconventional level of moral reasoning scale. The findings imply that intervention programs which access desired attitudes toward petrified wood and establish a credible threat of punishment will be the most effective in reducing petrified wood theft. / Master of Science

Mean overwintering survival of beetles was 20.5 ± 17.2% and 21.3 ± 3.6% for two consecutive years. Leaf litter was the preferred hibernaculum followed by straw. Successful overwintering was not compatible with water saturated soil. The mean supercooling point of C. rubiginosa in summer was -5.2 ± 1.5°C and was significantly different (P<0.05) from mean supercooling points of -9.3 ± 2.1°C in fall and -8.8 ± 2.3°C in winter. Females studied individually produced an average of 61.1 ± 29.9 oöthecae/female under field temperatures for the period of May 2 through August 13. Development from egg to adult under field temperatures required an average of 27.4 ± 1.7 days when reared on Canada thistle and 28.5 ± 2.1 days on an alternate host, musk thistle. Mortality due to incomplete development totaled 52.5% for immatures reared on Canada thistle and 79.0% for immatures reared on musk thistle. Mean survival from egg to adult of beetles maintained in field cages ranged from 18.3 ± 26.7%. Mean survival from early (1st - 3rd) instar to adult averaged 22% for beetles confined in field cages and 16% for beetles not confined in cages. Ingestion of leaves treated with the herbicides glyphosate and 2,4-D had neither a positive nor negative effect on amount of feeding by C. rubiginosa adults and did not appear to affect the longevity of adult beetles or the fecundity of females. / Master of Science

The oxygen stable isotopic compositions of minerals give valuable insight into the processes that govern many low- and high-temperature geologic systems. Much research has focused on the ability to analyze the fine-scale oxygen isotopic zonation in many minerals using techniques such as the laser probe and ion microprobe. In this study, I examine the feasibility of making high-resolution, nondestructive oxygen stable isotopic analyses with precision comparable to current microbeam techniques (i.e. <1‱). Five natural and two synthetic ¹⁸O doped calcites were analyzed using the laser Raman microprobe (LRM) to determine whether the intensity ratio of the C¹⁸O¹⁶O₂²⁻ and C¹⁶O₃²⁻ symmetric stretching bands can be directly related to isotopic compositions measured by ratio mass spectrometry (RMS). The calcites were cooled to near liquid nitrogen temperatures (78-90K) to remove the background from Mn²⁺ fluorescence and minimize the overlap of the symmetric stretching bands. At near liquid nitrogen temperatures, the presence of the C¹⁸O¹⁶O₂²⁻, C¹⁷O¹⁶O₂²⁻, and C¹⁶O₃²⁻ symmetric stretching bands are clearly visible. Initially, LRM isotopic values showed poor correlation with RMS isotopic values; however, the difference between LRM and RMS isotopic values was shown to be a linear function of the halfwidth of the symmetric stretching band. Corrected LRM isotopic values for all of the natural calcites agree with RMS values to within 3.8‱. Currently, the precision and accuracy of LRM isotopic measurements on calcite are several times poorer than for RMS measurements. The major limitation on the precision of the measurements is counting statistics. The strong Mn²⁺ fluorescence background of many natural calcites may also be a limiting factor in the precision of LRM isotopic measurements. However, the use of alternate laser excitation wavelengths and analysis at lower temperatures may allow the Mn²⁺ fluorescence background to be eliminated in these calcites. / Master of Science

The two dimensional fast Fourier transform (FFT) is a useful operation in many digital signal processing applications, but it is often avoided due to its large computational requirements. This thesis presents the implementation and performance figures for the fast Fourier transform on an FPGA-based custom computer. The computation of a 2-D FFT requires O(N²log₂N ) complex floating point arithmetic operations for an NxN image. By implementing the FFT algorithm on a custom computing machine (CCM) called Splash-2, a computation speed of at least 180 Mflops and a speed-up of 23 times over a SPARC-10 workstation is achieved. / Master of Science

Special methods were developed to model adhesively bonded joints in two- and three- dimensions using a simple finite element approach. Analysis of a two-dimensional bonded joint is performed by using plane frame elements to model the adherends, and a single plane elasticity-type element through the thickness of the adhesive bondline. A similar approach was developed for the analysis of bonded joints in three-dimensions such that the adherends were modelled by shear deformable plate elements, and the adhesive as a single solid element through the thickness. The degrees of freedom of both the plane elasticity ADH2D element and the solid ADH3D element are offset from their respective surfaces to the nodes of the adherend elements in each case, such that displacement continuity is achieved at element interfaces. The ADH3D-plate formulation can be used to analyze tapered adhesive layers, stepped laminated composite adherends, and thermal and moisture expansion effects in both the adhesive and adherends. A single lap shear joint was modelled in both two- and three-dimensions using the ADH2D-plane frame, and the ADH3D-plate configurations respectively. Adhesive stresses in two-dimensions converged to accepted closed-form solutions. Significant three-dimensional effects were observed in the ADH3D results, and possible explanations for this behavior were given. A typical crack-patch repair scenario was also modelled in two- and three-dimensions using the ADH3D formulation. The use of appropriate boundary and loading conditions for modelling such applications were discussed. Adhesive joints can be accurately and efficiently modelled using the ADH2D-plane frame, and the ADH3D-plate methods. In the process of selecting the most appropriate element to model the adherends in the three-dimensional ADH3D formulation, the phenomenon of shear locking in plate finite elements was examined and explained in terms of the presence of boundary layer-type solutions to the equations of shear deformable plate theory. To demonstrate this, the governing equations of Reissner plate theory were derived and reduced to independent equations expressed in terms of a displacement potential <i>φ</i> and a rotational stream function <i>ψ</i>. A plate finite element was derived using an interpolation of the displacement potential <i>φ</i>. Shear-locking was not observed when square plates with simply-supported and clamped edges were modelled using this finite element. A discussion on the actual cause of shear-locking and recommendations for future development and implementation of the concepts in this study were made. / Master of Science

Basinwide visual estimation techniques were used to inventory habitat and fish populations in three Shenandoah National Park watersheds having low, moderate, and high acid neutralizing capacity (ANC) to determine: a) types, distribution, and total amount of habitat; b) community structure and estimate fish density, condition, and production; and c) the influence of habitat verses ANC on fish communities. Fish species richness ranged from 3 to 7 species, and increased with ANC. Brook char (Salvelinus fontinalis) and blacknose dace (Rhinichthys atratulus) were the only two species common to all watersheds. Both species used pools more than riffles (P<0.05). Brook char density in pools increased with ANC (P<0.05), ranging from 13.1 to 40.1 fish/100m². Blacknose dace density in pools was similar between the low ANC stream (23.2 fish/100m²) and the high ANC stream (22 fish/100m²) in spring 1993, but eventually became higher in the low ANC stream (P<0.05). Annual production by brook char ranged from 23.7 to 26.4 kg/ha in the low ANC stream, from 27.7 to 39.9 in the moderate ANC stream, and from 48 to 77 kg/ha in the high ANC stream. Annual production by blacknose dace generally increased with ANC, ranging more narrowly from 0.3 to 2.4 kg/ha. Condition factor and mean weight of both species were lowest in the low ANC stream (P<0.05 for each). The effect of water acidification on movement of brook char and blacknose dace was studied in the paired channels of a laboratory stream. The objectives of this experiment were to determine the ability of the fish a) to avoid depressions in pH, and b) to use an alkaline microhabitat refuge during reductions. Both brook char and blacknose dace avoided an acidified channel and sought refuge in a refuge (P<0.05). / Master of Science

The purpose of this study was to determine what effects introductory Total Quality Management (TQM) training, has on the differences in paired supervisor and subordinate ratings of supervisor task performance. I hypothesized the differences in ratings would decrease after the training. This research was a quasi-experimental field study. My unit of analysis was established supervisor and subordinate pairs. Two established American university administrative offices provided the experimental (10 pairs) and control (6 pairs) groups. A Solomon Four Group Design was used; subjects were randomly assigned to pre-test; all other assignment was voluntary. There was no subject duplication. Each supervisor had a single subordinate and each subordinate had a single supervisor. The treatment was a series of Deming Library videotapes followed by an open discussion over the course of three months. Experimental group participation in the treatment was mandatory but the response to the survey was voluntary. The measurement instrument was broad: 118 task-related questions were clustered into 21 subscales based on previous work by the U.S. Merit Systems Protection Board. A seven-point scale was used. A separate random sample of people associated with the university identified instrument subscales of greatest importance to the supervisor-subordinate relationship. The subscales were leadership, communication, and supervision. The primary analysis was a series of independent t-tests, comparing differences between paired differences. As anticipated due to the small sample size, few significant differences were observed. At post-test, a comparison of the supervisor and subordinate views differed. Experimental group subordinates tended to have more favorable views of their supervisors’ performance while control group subordinates’ ratings tended to be less positive than the supervisor's views. The experimental group seems to have weathered the budget cuts with a more positive outlook. / Master of Science

Advanced power generation systems require regenerable sorbents capable of removing sulfur from coal gas to low ppm levels. Zinc titanate (Zn₂TiO₄) based sorbents are currently the leading sorbent for sulfur removal at high temperatures (~700°C). Zinc titanate sorbents are still in need of development to address the problems of zinc vaporization, physical degradation, and low attrition resistance during the many sulfidation and regeneration cycles. Based on the principles of phase equilibria and crystal chemistry, the approach being investigated in this project to eliminate these problems with zinc titanate sorbents is incorporation of various cations into the Zn₂TiO₄ spinel structure. The study involved synthesis, characterization, and evaluation of the modified Zn₂TiO₄ sorbents at the labscale size. Samples were characterized by XRD and TGA under simulated coal gas and an H₂S gas mixture to 1) determine the solid solubility of five cations (Ni, Cr, Al, Mg, and Cu) in the Zn₂TiO₄ lattice and 2) determine the effect of the cations on the sulfur removal performance of Zn₂TiO₄ sorbents. The five cations selected were incorporated into the Zn₂TiO₄ lattice in significant amounts, up to ~35 mole percent for Ni, Cr, Cu, and Al and ~25 mole percent for Mg at 1100°C. Based on the high chemical reactivity and durability observed in preliminary screening, Cr-incorporated Zn₂TiO₄ sorbents, with compositions ranging from 3 to 35 mole percent Cr, were selected for further testing and characterization, including XRD, TGA, crush strength, and EDX. No correlation was found between Cr concentration and porosity, crush strength, and weight gain (sulfur removal). Cr-incorporated sorbents reduced Zn losses from 16.4% with 22.2 mole percent to 1.2% loss with 26.1 mole percent Cr. / Master of Science