PP1-Arch was verified as a protein phosphatase by both acid molybdate extraction and thin layer electrophoresis. Soluble fraction was prepared from <i>Sulfolobus solfataricus</i>, from which PP1-Arch was purified over 1OOO-fold by DE-52 ion-exchange, hydroxyapatite, gel filtration (G- 100), and Mono Q FPLC chromatography. PP1-Arch was identified from the final purified sample by renaturation on an SDS-polyacrylamide gel. The molecular size of PP1-Arch was determined by both gel filtration chromatography and SDS-PAGE as 28 kDa and 33 kDa, respectively, which suggests that PP1-Arch is a monomer. PP1-Arch was found stable at temperatures as high as 90°C. Activation constants for the divalent metal ions Mn²⁺ and Ni²⁺, and the K_m for phosphocasein were determined. Myosin light chain was found to be a substrate for PP1-Arch <i>in vitro</i>. EDTA, Cu²⁺, Zn²⁺, P_i' and PP_i were shown to be inhibitors of PP1-Arch, while many compounds known to affect eukaryotic protein phosphatase activities were found to be without noticeable effect. N-terminal and an internal peptide sequence of the enzyme were obtained. The gene for PP1-Arch was cloned by a combination of "touchdown" PCR and conventional cloning techniques. The PP1-Arch gene was sequenced on both strands, and the sequence was compared with ones from eukaryotes and bacteriophage λ. The sequence homology between PP1-Arch and PP1/PP2A/PP2B suggests that they belongs to the same genetic family. A recombinant plasmid which was derived from pT7-7 was constructed for expression of PP1-Arch. The PP1-Arch gene was expressed in <i>E. coli</i> and the activity of the expressed enzyme was tested and shown to be divalent metal ion-dependent. Formation of inclusion bodies of expressed PP1-Arch was demonstrated. / Ph. D.

This dissertation addresses several key issues related to high-frequency soft-switching PWM three-phase power converters. These are: 1. Analysis, synthesis, and design of three-phase soft-switching PWM power converter topologies 2. Design of input EMI filters for three-phase converters 3. Design of microprocessor controllers for three-phase converters. An analysis of existing soft-switching PWM techniques is performed, and two generalized soft-switching PWM converter circuit representations are derived. Based on these representations and common topological properties of three-phase and dc-dc PWM converters, two new procedures for synthesis of three-phase soft-switching PWM converters are derived. The two procedures are used to synthesize five new three-phase soft-switching PWM converter topologies suitable for wide range of applications. A digital signal processor-based controller implementation example is presented. It demonstrates the feasibility of producing versatile, high performance, reliable, low-cost digital controllers for soft-switching PWM three-phase power converters operating at high switching frequencies. A new approach to the design of input filters for ac power electronic circuits is presented here. This approach is based on the application of a vast body of knowledge about passive L-C filters that has existed for many years, but has not been used in power electronics. New passive and active filter pole damping schemes are applied to high-order elliptic filters, resulting in significant filter size reduction compared to the standard filter designs. / Ph. D.

The purpose of the current study was to investigate whether the separate contributors to procedural and distributive justice also affected the other form of justice. Previous research investigating these cross over effects of justice contributors had not examined inputs in addition to outcomes and procedures, and had typically assumed outcome level to be equivalent to the equitableness of outcomes. Subjects were 120 undergraduate psychology students. Outcomes, inputs and procedures were manipulated in a laboratory experiment in order to assess their independent and combined effects on distributive and procedural justice perceptions. In contrast to past research, the current study found a weak and inconsistent effect of procedures on distributive justice perceptions. Outcome level had a strong effect on both procedural and distributive justice perceptions. In addition, outcome fairness was found to effect procedural justice perceptions. When procedures were fair, the equitableness of outcomes influenced distributive justice ratings. When procedures were unfair, however, the equitableness of outcomes did not influence distributive justice judgements. Implications for procedural justice conceptualizations, equity theory and organizations are discussed. / Ph. D.

The current study was designed to test the utility of a multivariate rnodel to predict health care utilization, absenteeism, and achievement functioning. The model included demographic factors and measures of physical health status as well as psychosocial predictors (child mental health, child developmental status, and family functioning). Four dependent variables were used including encounters with the health clinic at schoot primary care utilization, absenteeism (partial and whole days for the current year), and achievement status. Multiple regression analyses were performed to determine whether psychosocial factors would account for significant portions of the variance beyond that predicted by health and demographic factors. Results revealed that psychosocial predictors contributed significantly to the prediction of school and primary care health encounters. Predictors of absenteeism and achievement also included contributions of psychosocial factors although they were less robust. / Ph. D.

Clostridium difficile is a Gram positive, anaerobic bacterium which produces two potent protein toxins, A and B. The genes for toxins A and B have been previously cloned and sequenced and lie within 1.4 kb of each other. Upstream and downstream boundaries between sequences shared by both toxigenic and nontoxigenic strains and those sequences which are unique to toxigenic strains were established. A toxigenic element was defined in C. difficile strain 10463 which is 19.6 kb in length and is comprised of five open reading frames, including the toxin A and B genes. One of these open reading frames is previously unidentified and is located upstream of toxin B. Products of Polymerase Chain Reaction (PCR) amplification of three regions in the toxigenic element: the upstream boundary, the downstream boundary, and the region between the toxin A and B genes, were all identical in length in six toxigenic strains, indicating that the toxigenic element is conserved among these strains. A short fragment unique to nontoxigenic strains and occupying the same position on the chromosome as the toxigenic element was identified. peR products of this region were identical in length in three nontoxigenic strains. Transcriptional analyses were undertaken using probes to each of the five open reading frames in the toxigenic element. Transcripts were detected for four of the open reading frames which are contiguous and transcribed in the same direction. In addition, a very large transcript, corresponding to the length of the four open reading frames and processing intermediates were detected, indicating that the toxin genes are cotranscribed. A promoter region and processing sites were identified. Sizes were determined for each of the individual transcripts which correspond well with the sizes of the open reading frames. Six toxigenic strains which vary considerably in toxin production were selected for analysis to determine whether DNA sequence variation could account for the observed differences in toxin production. DNA restriction fragment length polymorphisms were examined, toxin-specific transcripts were analyzed, and sequences of regulatory regions were determined and compared. Whereas quantitative differences in toxin-specific transcripts were found among the toxigenic strains, the remaining analyses showed that DNA sequences were conserved among these strains. / Ph. D.

Analysis of the flow field in a laser engine represents a difficult computational problem involving combinations of complex physical and gas-dynamical processes. Following a brief discussion of these processes a calculation procedure using primitive variables formulation on a non-staggered grid system is introduced. Based on this procedure, a Pressure Based Navier-Stokes Solver (PBNS) is developed using a generalized curvilinear coordinate system. The solver is first tested in application to a subsonic compressible flow over an insulated flat plate and to a flow in an axisymmetric converging-diverging nozzle. Next, the PBNS code is used to analyze the flowfield and performance of a laser thruster. The physical/numerical model includes the geometric ray tracing for the laser beam, beam power absorption, plasma radiation losses, and plasma thermophysical and optical properties. Equilibrium hydrogen is used as a flowing gas and its properties are calculated using the Hydrogen Properties Calculation (HPC) based on the the methods of statistical thermodynamics. Two thruster configurations, two laser types (CO₂ and iodide), various laser power levels, and various injection conditions are tested. The results of these tests include the temperature, pressure, velocity and Mach number contours, as well as table of the laser beam power absorbed, radiation losses to the thruster walls, thrust level and specific impulse. The maximum specific impulse obtained in these tests is 1537 sec for a CO₂ laser thruster and 827 sec for an iodide laser thruster. Up to 100% power absorption can be achieved, however, the radiation losses from the hot plasma are quite high disallowing a full conversion of the absorbed power into the thermal energy of the propellant. The PBNS code can be used to study the effects of various design parameters on the performance of a laser thruster and provide guidelines for the preliminary design of a laser engine. / Ph. D.

Tractor rollover protective structures (ROPS) have been found effective for protecting people from serious injury or death when tractors overturn. Extension agricultural engineers and safety specialists are promoting the retrofitting of ROPS on pre-ROPS tractors. ROPS, meeting to ASAE standards, are available for these older tractors. However, effective protection not only depends on the proper design of ROPS, but also on the structural adequacy of the rear axle housing on which ROPS are mounted, and on the means of attachment. Adequate information on the strength of pre-ROPS tractor axle housings 1s still lacking. A research project to develop a technique for determining the suitability of tractor axle housing when adding ROPS to pre-ROPS tractors was initiated by the Biological Systems Engineering Department at Virginia Tech in 1991. A finite element procedure was developed for the rear axle assembly of John Deere 720 tractor and models were built with a commercial software ANSYS Engineering Analysis Program. The models were employed to analyze the stress distributions under the static, crush, and dynamic ROPS tests recommended by ASAE. Laboratory tests were conducted to validate the models, and to assess the suitability of the shell element model and the general solid element model. Results of the analyses show that the rear axle housing assembly of John Deere 720 tractor has a safe margin for the static and crush tests under the condition of this study. Results also show that can support the dynamic test if the ROPS is mounted with a larger contact area with the axle assembly. The study concluded that the finite element procedure is effective to assess the strength adequacy of tractor axle housing and to determine the feasibility of retrofitting ROPS on pre-ROPS tractors. / Ph. D.

A product portfolio approach is used in this dissertation to develop a model permitting capital budgeting to be modeled interactively with aggregate production planning, in light of market supply and demand functions. Primary emphasis is on the maximization of profit, but other goals are also addressed. These are maximization of the rate of return, maximization of market share, and minimization of the cost of excess capacity. A linear mixed integer programming model is developed for each of these objectives. Then, a single goal programming model that combines all four objectives is formulated. Costs are not allocated to products. Accordingly, the notion of cash flows per product (or per project) is not used. Instead, cost is incurred as a result of the demand that a product portfolio places on resources. All costs are considered to be incurred in the acquisition and utilization (in the form of activities) of resources. Four distinct levels of activities are considered: unit, batch, product sustaining, and facility sustaining. The demand for each resource is aggregated over all levels of variability and over all the products in the product portfolio. The direct cash outflow or inflow as a result of changing resource capacity is continuously traded off against the eventual cost or benefit of changing the capacity (in the form of changed revenues and as a function of both time and market supply and demand). Capital structure and capital investment decisions are considered simultaneously for a given set of assumptions. Different sources of funds are utilized for different costs of capital. Lending and borrowing are simultaneously incorporated without the solutions becoming inconsistent due to incorrect or inappropriate discount factors. This is mainly attributable to the fact that the organization, as a single entity that manufactures a product portfolio, demands capital, and invests excess funds. The net present value of the organization (not of projects or products) is maximized. Also, the output of each project is modeled specifically. This alleviates the practical problem of fractional acceptance of projects. Variable market supply and demand functions are also included and modeled explicitly. Finally, it is shown that the developed model contains several elements of aggregate production planning. The main conclusions from this research are: 1) Better capital budgeting results can be obtained if costs are not allocated to projects (or products) when resources are shared among different projects or products; 2) Financing and investment decisions can be made interactively (with the developed model) without the solutions becoming inconsistent due to unknown discount rates; 3) Resource acquisition and resource consumption should be modeled explicitly in capital budgeting; and 4) The model yields an improvement over existing capital budgeting techniques for a given set of assumptions. Some recommendations are presented for further research to extend these conclusions. / Ph. D.

Improving protein quality and identifying superior inbreds and hybrids are significant challenges in commercial maize breeding programs. These two problems were addressed in separate studies on inbreds and hybrids from a complete diallel cross of 12 elite proprietary inbred lines of maize evaluated in field trials in two locations for two years. One of the inbreds (WI) was a novel source of high quality protein obtained from Wilson Seeds, Inc. in Harlan, Iowa. In the first study, diallel analyses were used to study combining ability and types of gene action important in the inheritance of protein content, grain yield, grain moisture at harvest, time to silk, kernel hardness, and density. General combining ability (GCA) and specific combining ability (SCA) effects were highly significant for all traits indicating presence of both additive and non-additive effects, respectively. Reciprocal effects (REe), often assumed to be absent in maize diallel studies, were significant for grain yield and protein concentration, suggesting that choice of female parent may be important for these traits. Ratios expressing the relative importance of GCA and SCA indicated that protein concentration is controlled primarily by additive gene action. In the second study, restriction fragment length polymorphism (RFLP) data were obtained for the 12 inbreds using 42 genomic clones each with four restriction enzymes. Modified Roger's distances were calculated and used in cluster analyses for heterotic grouping of the inbreds. Two measures of level of heterozygosity and hybrid value were evaluated as means of predicting Fl performance of hybrids in the complete diallel set of hybrids and in groups of hybrids representing crosses between and within heterotic groups. Results from this study confirm those of previous investigations with respect to prediction of hybrid performance when comparable groupings of crosses between related and unrelated lines were evaluated. This study further indicates that RFLPs may also be useful for prediction of hybrid performance in situations typical of early generations of many maize breeding programs where recombinant inbreds are testcrossed to a common tester inbred. / Ph. D.

In our previous studies, uncross-linked large diameter cellulose beads were optimized for solids content, bead size, pressure-flow limits, molecular accessibility and performance as an immunosorbent. Here, anion exchange (DEAE) cellulose beads were derivatized by two different procedures (defined as A and B) and the changes in bead morphology were correlated with transport and sorption kinetics. The kinetic characteristics clearly defined a minimum of two different types of protein binding site architecture. DEAE cellulose beads exhibited molecular exclusion of BSA near the edge of the bead in contrast to greater permeability seen in underivatized beads. Thus, accessible BSA binding sites are present only on the surface of the derivatized beads. DEAE cellulose beads derivatized by procedure B gave higher density of DEAE ligand as compared to beads derivatized by procedure A, as well as higher static and dynamic capacity for BSA. Even though DEAE cellulose beads (DP 2070, 450 μm diameter derivatized by procedure B) have lower small ion capacity than DEAE cross-linked agarose beads, as well as 1/4 the surface area, they exhibit equivalent binding capacity for BSA per volume of support. Thus, DEAE cellulose beads possess more sites per surface area as well as have lower ligand density per BSA site. Furthermore, BSA adsorption sites on DEAE cellulose beads derivatized by procedure B exhibit slow binding kinetics as compared to those derivatized by procedure A and also compared to DEAE crosslinked agarose beads. Thus, the rate limiting step for the adsorption of BSA on DEAE cellulose beads was not diffusion as suggested by the large diameter of the bead. Feasibility studies were performed for process scale applications to fixed and expanded bed anion exchange purification. The large diameter DEAE cellulose beads of this study maybe useful for process scale anion exchange as evident from purification of immunoglobulins from hybridoma cell culture in fixed bed. The balance of large diameter and density of these DEAE cellulose beads enable stable expanded bed purification of proteins such as recombinant human protein C from transgenic porcine whey. / Ph. D.