We describe an adaptive filter system that is able to switch between adaptive filter algorithms in order to produce fast convergence and low mean Square error (MSE). The switching system employs two adaptive filters for two different tasks; one is intended to yield fast convergence, and logically called the "fast convergence structure", while the other is intended to give small MSE, and called the "low MSE structure". The switching from one algorithm to the other is determined by the state of the system. For example, switching from the "fast convergence structure” to the “low MSE structure” happens if the former has reached its steady state according to some pre-defined criterion, while switching from the “low MSE structure" to the "fast convergence structure” happens if the former starts diverging according to some pre-defined criterion. We define here that an algorithm has reached its steady state if the average of the square of its output error is small and approximately constant for several iterations. After an algorithm has reached its steady state, not much additional error reduction can be obtained from it, so that there is no payoff using "the fast convergence structure”, which is usually more computation intensive than the "low MSE structure”. In this situation it would be better to use the Least Mean Squares (LMS) algorithm as the “low MSE structure" because of its simplicity and its numerical robustness. Experiments using the recursive-least-squares-lattice (RLSL) algorithm together with the LMS algorithm, the fast-transversal-filter (FTF) algorithm together with the LMS algorithm, and the gradient-adaptive-lattice (GAL) algorithm together with the LMS algorithm for a system identification application, in particular for echo cancellation, show the expected result of providing faster convergence and lower mean square error than would be possible with a single algorithm. The switching system demonstrates other important results: it can avoid the numerical instability of some algorithms, such as the RLSL and FTF algorithms, without adding any additional computations; it is able to handle a change in the unknown system, as long as it settles, without suffering a slow convergence rate caused by an incorrect initial condition; it is able to handle a change of the observation noise without facing a divergence problem; and it is able to produce an optimum result even for i/l-conditioned input signals, i.e. the ratio of the maximum and minimum eigenvalue of the auto-correlation of the signal is high. When switching to the “low MSE structure" we also apply a computationally reduced order technique, in which only the values of the impulse response that are greater than some threshold are used for computation. This technique is applied to the switching structure of the recursive-least-squareslattice algorithm together with the LMS algorithm and exhibits fast convergence and low MSE even for ill-conditioned input signals. For the white Gaussian noise input, on the other hand, this technique yields a somewhat larger mean square error. / Master of Science

The Base Test Method has been shown to be a viable means to predict the capacity of gravity loaded standing seam roof systems. However, for uplift loaded standing seam roof systems, the base test method is still being investigated. Previous tests conducted at Virginia Tech in 1990 and 1991 have led to the belief that scatter is inherent in the results of the base test method for uplift. It was found in this research that scatter does exist in the base test method for uplift loading. Tests were also conducted to determine if purlin size and span could be eliminated from a manufacturer's testing program. It could not be shown in this research that a relationship exists between the percentage of through-fastened capacity, 'R', and the constrained bending capacity of a purlin. / Master of Science

A modified eddy correlation technique is applied to data obtained from a field experiment conducted in the Bight of Abaco during the Spring of 1990. The experiment yielded 3408 30 minute records of turbulent wind data sampled at 5 Hz. Wind stress estimates were successfully calculated for 3182 of these. The modification to the eddy correlation technique involved extending the cospectra to higher frequencies than were measured by the propeller anemometer to obtain a better estimate to the <i>uw</i> covariance. This resulted in a small increase to the <i>uw</i> covariance, and hence the drag coefficient. / Master of Science

The PSD permitting of new sources has drawn increasing public and governmental opposition. Much of the opposition is due to heightened environmental awareness. The issues are further compounded by inherent obstacles resulting from the PSD regulations and guidelines. A chronological review of the process used by Old Dominion Electric Cooperative to obtain a PSD permit is presented to demonstrate the major PSD permitting issues facing new developers. The issues include the appropriate methodology for Best Available Control Technology determination, protocol to be used for air quality analysis, the best way to handle the Federal Land Manager's current concerns, and public participation. The present case study is analyzed to develop suggestions that would improve the current PSD process without jeopardizing the integrity of the program. Additionally, recommendations are made for developers who may encounter similar obstacles. / Master of Science

The proposed “natural crutch” provides ambulation assistance without the problems associated with arm-supported crutches. The new crutch has an exoskeleton structure which surrounds the disabled leg and carries the user's body weight while walking and standing. The load is transferred from the exoskeleton to the trunk of the body through a harness which lifts the pelvis on the impaired side. The intention is to simulate the internal skeleton’s natural load distribution while compensating for the impairment. The crutch design is presented as a unique concept with an outlined proposal for a marketable prototype. Results of a full scale simulator test are used to determine areas requiring further development. All aspects of design attempt to include both engineering and industrial design perspectives. / Master of Science

Register allocation is critical to processor performance. Registers are the fastest storage system available to a processor. The more capable a register set's organization is at maintaining process context, the fewer the number of memory accesses the processor will need to make. Overlapping register windows have better context maintenance capabilities than single register set organizations, but overlapping register windows also show significant performance degradation if program behavior causes the register window store to overflow. Program behavior makes window overflow of simple overlapping register window organizations unavoidable. Attempts to minimize the impact of overflow by increasing the size of the register store negatively impact register access time, increases device count, and increases context switch latency. The combination of a transparent spill and refill mechanism and a small register store, allows the store to perform like a much larger store, but does not negatively impact register cycle time, and it decreases context switch latency. Transparent register spilling and refilling can be accomplished by the inclusion of a set of simple state machines, and dedicated register and memory ports. The transparent spill/refill mechanism's external port interfaces very well with established peripheral processing capabilities on many multi-processor architectures. The inclusion of an instruction repetition capability can facilitate global register storage and retrieval, and can decrease context switch latency. Register performance can be further enhanced by partitioning the register set into data typed. register groups. Register partitioning allows a high degree of parallelism, without necessitating the inclusion of register set with high port counts and register access conflicts. Partitioned register sets can the spatially proximate to processing units whose functionality is optimized for operations on specific data types. A remote instruction pointer with a partitioned code address register set and processing capability can decrease branch latency, improve call/return performance, and simplify general case return address maintenance. A partitioned, transparently spilled/refilled register organization minimizes explicit register storing and retrieving, supports the creation of large register-based working sets, and facilitates a simple parallel processing paradigm that allows a high degree sub processing unit independence. / Ph. D.

The polyhydroxy character of lignocellulosics and their natural abundance make them good candidates for the manufacture of polyurethanes. The cure characteristics of hydroxypropyl-cellulose and hydroxypropyl lignin (HPC and HPL, respectively) with polymeric methylene diphenyl diisocyanate (MDI) was studied via dynamic mechanical thermal analysis (DMTA). HPC/MDI and HPL/MDI resins flow at 30°C and proceed to cure at 50°C. The latter has excellent thermal stability over the former. Crosslinking of HPL and HPC with MDI follow an nth order kinetics, with an order of reaction of 2 and an apparent activation energy in the range of 12.9 kcal/mol - 14.7 kcal/mol. The rate of cure with time is higher in HPL-based polymers than HPC-based ones at the initial stage of cure; the difference vanishes at later stages. This demonstrates that the hydroxyl groups in HPC are less accessible to the NCO groups, and that cure rate might be dependent on diffusion limitations at later stages. Degree of cure, under all cure schedules, follows a parallel trend, and has to do with the fact that the hydroxyl groups of HPC are less accessible to isocyanate. Both HPL and HPC react with MDI at a reduced rate in comparison to a synthetic polyol: caprolactone triol. Time-glass transition temperature superposition was used to calculate times to vitrification of the HPL-based polymers, and is presented in a TTT cure diagram. This bio-based polymer displays the s-shaped vitrification pattern characteristics of thermosets. A similar approach did not work with HPC-based polymers. HPC- and HPL-based polymers did not display damping transitions, in isothermal cure, typical of gelation and vitrification. As the isocyanate to hydroxyl ratio (NCO:OH) increased, the glass transition temperature of the polymers increased, and the transition amplitude and width decreased and increased, respectively. In practical terms, this study illustrates that it is advantageous to use a) to use high isocyanate to hydroxyl ratios in order to produce polyurethanes which retain desirable damping behavior over a wider range of temperature. b) to use HPC/MDI resins in those situations where retention of stiffness at temperatures below 230° is required. c) to use HPL where rapid cure is desired. The study also reveals that the relative reactivity of water, HPL and HPC with isocyanate takes the form water > HPL > HPC. / Master of Science

Defects formed in solid rocket propellant during manufacturing, transportation, storage, and assembly can lead to alterations in the thrust time profiles and possibly catastrophic failure of the entire rocket. In order to determine the effects of temperature, loading rate, and thickness on this particulate composite, tests were conducted at three temperatures and two loading rates. Both uncracked and edge cracked "biaxial" specimens were produced from solid rocket propellant. The stress relaxation modulus and stress-strain data were obtained from load curves formed during "biaxial" tension tests. Near crack tip displacements and strains were calculated from photographs taken of a surface grating on the pre-cracked specimens during crack propagation. The effect of thickness, temperature, and loading rate on the stress intensity factor was also studied. Finally, by applying continuum theory the displacement singularity was determined at different stages of crack growth. From the stress strain data, it was found that temperature had a greater influence on behavior than loading rate over the ranges studied. The crack growth in the composite material consists of a series of crack opening, crack blunting, and crack growth/resharpening stages which are highly nonlinear. However, the thick specimen at low temperature did not follow this crack growth mechanism. At -65°F the thick specimen developed transverse constraints which caused a brittle fracture to occur when the specimen was loaded. Determination of the displacement singularity order for the sharp cracks was found to be consistent with the theoretical results predicted by Benthem. / Master of Science

Zinc gallate and Mn²⁺-activated zinc gallate were identified as potential low-voltage cathodoluminescent phosphors for use in vacuum fluorescent displays. The stability of these oxide phosphors in high-vacuum and absence of corrosive gas emission under electron bombardment, offer advantages over commonly used sulfide phosphors. A low-voltage cathodoluminescence spectrophotometer was developed for phosphor characterization. Sample brightness was measured as a function of anode voltage (10-300 VDC). The effects of activator concentration, phosphor layer thickness, deposition process, and internal pressure were examined. From photoluminescence measurements, absorption and emission centers were identified, the role of composition in the luminescence process explained, and host-to-activator, non-radiative energy transfer identified for ZnGa₂0₄:Mn²⁺. Samples of the general composition Zn_1-xMn_xGa₂O₄, with x ranging from 0 to 0.03, were synthesized by solid-state reaction techniques using oxide precursors fired in air, followed by reduction firing in 98%N₂, 2%H₂. The phase-pure ZnGa₂O₄ spinel structure of all the compositions was characterized by X-ray diffraction. / Master of Science

A special wafer bonding method called the Silicon Direct Bonding technique is used to study the bonding of unipolar (n-type, <100> oriented) silicon wafers. The primary objective of this thesis project is to find an optimum surface preparation technique for subsequent silicon wafer bonding. Wafer cleaning and treatment methods are investigated to understand the correlation between a high quality wafer surface and the resulting high quality electrical conduction at the interface. Accordingly, in this project, a preference for hydrophobic (less polar Si-OH surface) wafers is given to ensure a minimized amount of oxide layer on the surface. Several key factors that govern the quality of the wafer surfaces, such as the degree of hydrophobicity, HF etching time, composition of HF etching solution and Dr water rinse, are examined with ellipsometric and XPS measurements. An HF etching followed by a sputter etching has been selected to pre-treat the wafer surfaces for bonding. A maximum allowable air exposure time (35 second) is also found which would allow bonding without significant re-growth of the oxide layer. Bonding is performed under vacuum with a special mechanical fixture and the resulting structures from a subsequent heat treatment process are examined with crack propagation testing. Bond strength after annealing is sufficient to withstand a pull test, however, with a 3 point bend testing, the crack propagated horizontally at the interface. / Master of Science