This research presents an analysis of the improvement in the performance of a class of fault tolerant multistage interconnection networks. In the network discussed here, fault tolerance is achieved by providing multiple redundant paths between the source and destination. The extra paths are obtained by providing redundant links between switching elements within a stave (intra-stage links), thereby increasing the switching element complexity. The techniques used in the construction of this network, its properties, advantages, and disadvantages are discussed. While early studies focused their effort in analyzing the fault tolerant characteristics of the network and the performance in a circuit switched environment, this investigation complements the previous work by examining fie performance of a packet switched network. The reasons for the choice of the architecture that include factors like hardware complexity, cost and simplicity of control algorithm are analyzed. The study concentrates on improving the run-time performance of the fault tolerant network. by using these multiple paths not only in the presence of a fault, but also in a fault-free environment. The throughput of the packet switched network in the presence of a fault, congestion and when fault free are analyzed. A description of the investigation, assumptions and factors used for the study, a cost analysis, and the results of the simulation analyses is included. / Master of Science

The design and development of a system for disinfecting medical waste at the site of origin is presented. Investigation of the current commercial systems that accomplish this task shows that they all expose the waste to physical conditions that are harmful to all forms of life. Further, most are very expensive to install and to operate. A recently developed biochemical process promises to effectively inactivate harmful pathogenic organisms economically and without the danger of extreme heat or poisonous chemicals. The biochemical process is not yet fully developed. Nonetheless, the development of a marketable system to take advantage of this technology has been initiated. The motivation for developing this technology and the particular system that will employ it is presented. A general overview of the system and components is presented. Previous and suggested future testing strategies are explained. Component interactions and process control are described. / Master of Science

Pyruvate:ferredoxin oxidoreductase activity was demonstrated in microaerophilic members of Campylobacter, Wolinella, and Helicobacter. Arcobacter cryaerophila (sic) and Arcobacter nitrofigilis, two aerobic species, lacked any detectable activity. Under anaerobic conditions crude extracts of Campylobacter, Helicobacter and Wolinella were capable of reducing the electron carriers benzyl viologen and metronidazole in the presence of pyruvate. Addition of Clostridium pasteurianum ferredoxin to the metronidazole-linked reaction enhanced metronidazole-reducing activity, suggesting that electron transport to artificial electron carriers is facilitated by ferredoxin. All species exhibited varying degrees of sensitivity to metronidazole (MIC = <0.8 to 25 µg/ml) except Arcobacter cryaerophila, which was resistant to > 100 µg/ml. This further supports the theory that these organisms possess ferredoxin-linked reactions. The presence of the oxygen-labile enzyme pyruvate:ferredoxin oxidoreductase may be related to the inability of these microaerophilic bacteria to grow in normal atmospheric levels of oxygen. Under aerobic conditions crude extracts of the organisms were also capable of reducing NAD in the presence of pyruvate. This might be accounted for by an NAD-linked pyruvate dehydrogenase; alternatively, it might be due to an enzymatic reduction of NAD by electrons from the reduced ferredoxin generated during the ferredoxin-linked pyruvate oxidoreductase reaction. / Master of Science

Cold acclimation entails changes in membrane composition, osmotic adjustment, alterations in the cell wall-plasma membrane interface, sugar deposition, and changes in cell wall proteins. There is evidence that a rigid cell wall may be necessary for cold acclimation. Difficulties arise in studying plant material acclimated by exposure to low temperatures, because extraneous changes in the plant material occur that are unrelated to the development of cold tolerance. In order to determine whether cell wall changes are necessary for acclimation, peas were acclimated at warm temperature (26°C) by the application of exogenous ABA, desiccation, light exposure, and an experimental cryoprotectant (GLK 8908). Electrolyte leakage, elastic and plastic bend angles, and stem elongation were used to evaluate freezing injury, cell wall rigidity, and growth, respectively. The role of extensin, a structural hydroxyproline-rich glycoprotein suspected of being involved in cell changes during acclimation, was examined using the hydroxylation inhibitor ⍺,⍺’-dipyridyl. Exogenous ABA application and drought stress decreased freezing injury by approximately a 10% -6°C compared to controls. In one experiment light was found to be more effective than ABA at acclimating peas at warm temperatures. Foliar application of GLK 8908 decreased freezing injury (30% at -6°C). Stem bendability was not correlated with freezing resistance. ABA treated peas grown in the dark had reduced growth rates and increased stem rigidity, but exhibited greater injury at -6°C than untreated dark grown peas. Extensin content was not related to cold hardiness. Although acclimation of ’Alaska’ peas did occur at warm temperatures with various treatments the reductions in freezing injury were minor when compared to plants acclimated by exposure to low temperatures. GLK 8908 was also evaluated for its effects on ’Alaska’ pea survivability and yield. Peas treated with GLK 8908 (1 and 10% aqueous) and subjected to a -6.7°C freeze were found to have increased survival without significant changes in days to first flower, leaf surface area/plant, and yield/plant. / Master of Science

Medium solution N and electrical conductivity (EC), and foliar N levels were measured to determine criteria which signal the need for reapplication of a controlled-release fertilizer (CRF) in order to maximize growth of container-grown Ilex crenata Thunb. ‘Helleri’. Rooted cuttings of ‘Helleri’ holly were initially fertilized (7 Mar. 1991) with 8-9 month (18N-2.6P-9.9K) or 12-14 month (17N-3.1P-9.9K) Osmocote. A subset of 12 plants received a CRF reapplication (half rate) of the respective Osmocote formulation on 19 July, 2 Aug., or 16 Aug. In addition to reapplication treatments, 12 plants received a liquid fertilizer (LF) solution with each irrigation starting on 19 July. Canopy widths of plants receiving LF or CRF reapplication on 19 July and 2 Aug. were greater than those plants without supplemental fertilizer (control). At the time of the second reapplication date (2 Aug.), medium solution N and EC, and foliar N levels of control plants were = 40 mg liter⁻¹, 0.4 d m⁻¹ and 2.3%, respectively. Since growth of control plants was less than those of LF, CRF 19 July and 2 Aug. reapplications, medium solution N and EC, and foliar N values at the time of the second reapplication (2 Aug.) signaled the need for CRF reapplication or LF supplementation. There was a strong relationship between medium solution EC and N (NO₃ + NH₄) (r² = 0.91, P = 0.05), indicating that growers could use EC measurements to predict medium solution N concentration, as well as to assess the adequacy of the fertilizer regime. / Master of Science

Parametric design models were created to optimize cable sensitivities in a structural health-monitoring system. Experiments were also conducted to determine the sensitivity of a working system. The system consists of time domain reflectometry (TDR) signal processing equipment and specially designed sensing cables. The TDR equipment sends a high-frequency electric pulse (in the gigahertz range) along the sensing cable. Any change in electric impedance along the cable reflects some portion of the electric pulse back to the TDR equipment. The time delay, amplitude, and shape of the reflected pulse provides the means to respectively locate, determine the magnitude, and indicate the nature of the change in impedance. The change in the sensing cable impedance may be caused by cable elongation (change in resistance), shear deformation (change in capacitance), corrosion of the cable or the materials around the cable (change in inductance), etc. The sensing cables are an essential part of the health-monitoring system because the cable design parameters determine the cable impedance sensitivity to structural changes. By using parametric design models, the optimum cable parameters can be determined for specific cases and used to select or design an appropriate cable. Proof-of-concept and resolution experiments were also conducted to provide, respectively, verification and sensitivity of the system. / Master of Science

The wealth of experimental laser data and its high accuracy allow a rigorous comparison between a distributed model and distributed measurements that was not possible previously. The purpose of this work is to obtain and compare experimental and analytical distributed measurements for a simple structure excited by a point force (from a shaker) and by a distributed force (from a single or pair of piezoceramic actuators). The analytical results are obtained from the forced response analysis of a finite element model. A known piezoelectric actuator model that shows the actuation effects produced by symmetrically bonded actuators are line moments along the edges of the actuator and the effects produced by a single actuator are line moments and forces along its edges is used. A forced response analysis is performed on this model and the amplitude of the sinusoidal displacement at each prescribed point along the structure is found. Experimentally, a scanning laser vibrometer is used to measure the velocity along the structure due to the hannonic excitations. The vibration amplitudes are found from these results and compared to the modeling results. The predicted displacement shapes for the various excitations and boundary conditions resemble the measured shapes. The exact magnitudes of the predicted displacements along the beam do not corroborate the experimental results. The differences between the two are inconsistent over the length and width of the beam and over the excitation frequencies considered. Various modeling and experimental reasons are cited for this inaccuracy. The other moments from the edges of the piezoceramic actuators parallel to the length of the beam are found to influence the beam response even at low frequencies. The laser measures some torsional effects in the beam that are not predicted by the model. The measured responses show that at high excitation frequencies mode shapes indicative of a plate are excited in the beam by the piezoceramic actuators. / Master of Science

The radio interface greatly affects performance of wireless communication systems. Hard-wired communication links use transmission lines to connect communication terminals. The propagation characteristics of radio frequency signals on these transmission lines are well known. In wireless communication systems, however, the transmission line with a known impulse response is replaced by a radio channel with an impulse response that is constantly changing as the users roam throughout the coverage area. The varying impulse response is due to the multiple path propagation of the signals from the transmitter to the receiver. The design of emerging small cellular (commonly known as microcellular) wireless systems is limited by the multipath propagation characteristics of the channel. Once these propagation conditions are understood, systems may be designed more efficiently in terms of cell layout, interference reduction, and system performance. This thesis presents a technique for automated propagation prediction in outdoor microcellular radio channels using ray tracing. The basic method is to integrate site-specific environmental data with a geometrical optics model to trace the propagation of energy from the transmitter to the receiver. Software written in C++ is used to automatically trace rays that are reflected, transmitted, scattered, or diffracted as they propagate through the channel. The automated software uses AutoCAD® to maintain the site-specific building data incorporated into the model. Details of the building database, propagation model, and software implementation are included in this thesis. The accuracy of the model and its software implementation is tested against wide band measurements taken on the Virginia Tech campus. Results, included here, indicate that the received signal can be accurately predicted in both line-of-sight and obstructed microcell topographies. / Master of Science

Seventy-nine undergraduate students were interviewed in February, 1992, to determine attitudes about class attendance, class absences, and required attendance. Three hundred undergraduates were selected in a random sample; seventy-nine attended one of the six group interview sessions. The reasons students gave for skipping classes included being lazy or tired, dislike of the professor, material for the class was seen as unimportant, bad or nice weather, early morning or late afternoon and evening classes (too tired), or having other things to do (sometimes work for another class). Class size also was discussed as having impact on skipping. / Master of Arts

The activation energy of Douglas fir wood char gasified in carbon dioxide was determined. Activation energies were found for chars that had been pyrolyzed in nitrogen at 600, 750, and 900°C. A thermogravimetric analyzer provided the weight versus temperature data used to obtain the activation energies. The Coats-Redfern integral method of kinetic analysis was used to extract the activation energies from the data. This method can be used to obtain an activation energy from a single weight versus temperature trace for a constant heating rate. An overall apparent activation energy of 723 ± 60 kl/mole and a natural log of the pre exponential factor of 68.8 ± 6.2 was determined from the data collected for all three chars. The different char preparation temperatures did not appear to affect the activation energy. / Master of Science