No description available.

No description available.

No description available.

No description available.

A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of requirements for the degree of Master of Science. Johannesburg, 2014. / No abstract supplied.

Driven by Moore's Law, many computer architectures - ARM, x86, MIPS, PowerPC, SPARC - have evolved from 32-bit to 64-bit. To support existing applications, these have all kept support for a 32-bit compatibility mode. However, this comes at a cost in hardware complexity, power consumption and development time. Dynamic binary translation - recompiling binaries into the new instruction set at runtime - can be used instead of specific hardware for this purpose. While this approach has previously been used to assist architecture transition, these translators have all traded-off performance and transparency, a measure of how accurately they emulate the 32-bit environment. This thesis addresses ARM's transition from AArch32 to AArch64 through MAMBO-X64, a dynamic binary translator developed to support this transition. A range of novel optimizations were devised to improve translation performance while maintaining strict transparency. This follows a common theme of exploiting existing hardware features such as hardware return prediction, virtual memory and virtualization extensions to offset translation overheads. HyperMAMBO-X64 - a variant of MAMBO-X64 integrated in a hypervisor - was also developed to support system-level translation while remaining transparent to guest operating systems. Results demonstrate that the cost of binary translation is reduced, delivering performance competitive with the manufacturer's hardware. Performance in several benchmarks even exceeds that from the integrated compatibility mode. Thus MAMBO-X64 not only provides a means for architectural upgrade, but also an alternative to the expense of the legacy support currently employed.

In two parts, this thesis examined the perception of odour mixtures by humans. Several experiments were carried out and results shown. Part 1 established that peripheral interaction in the form of competition for transduction pathway resources is not of primary importance in determining the perceptual characteristics of odour mixtures. Other peripheral mechanisms, central factors and limitations in human cognitive processing would also appear to play an important role.Taking into account the temporal processing results found in Part 1, Part 2 sought to establish the capacity of humans to analyse odour mixtures in greater than two components in terms of the perceived order of components, and the identity of those components. Overall, the research demonstrated that two properties of simple binary odour mixtures, namely the temporal order and perceived strengths of components, are not primarily characterised by the transduction pathway those components putatively operate in human ORNs. / Doctor of Philosophy (PhD)

Global phase diagrams of binary systems in pressure-temperature (PT) space are very useful. In this project the techniques to automatically develop global phase diagrams in PT space were created. The codes to compute different components of a global phase diagram in PT space were developed. These codes were then successfully incorporated into a single functional program. <p>To generate the binary PT phase diagram, the overall composition was varied from pure component 2, the least volatile component (LVC) to pure component 1, the most volatile component (MVC). The step size for changing mole fraction was varied in the calculation of different parts of a global phase diagram. When the points near the joining points between different parts were computed, the step size was set to a rather small value. The step size was then increased to twice of the last value for each subsequent point computed. When the MVC mole fraction was approaching one, the step size was set to a small value to obtain enough points needed to minimize the chances of missing important phenomena. <p>The techniques to set initial guesses for evaluation of different components of a global phase diagram were discussed. The code performance, including the number of iterations for different convergence criteria and the sensitivity of the algorithm were presented. Using the code developed, phase diagrams of type I, type II, type III and type V were generated using representative binary systems from the petroleum processing field. <p>The boundary states between different types of phase behaviour were also explored. It was observed that with the increase of the binary interaction parameters, the phase behaviour of the ethane + ethanol binary system changes from type I to type II to type III while the methane + n-hexane binary system changes from type V to type III. These conclusions matched the results of van Konynenburg and Scott (1980). It was also concluded that with the increase of the binary interaction parameter for a binary system, the system showed a trend to exhibit more liquid-liquid immiscibility.

Global phase diagrams of binary systems in pressure-temperature (PT) space are very useful. In this project the techniques to automatically develop global phase diagrams in PT space were created. The codes to compute different components of a global phase diagram in PT space were developed. These codes were then successfully incorporated into a single functional program. <p>To generate the binary PT phase diagram, the overall composition was varied from pure component 2, the least volatile component (LVC) to pure component 1, the most volatile component (MVC). The step size for changing mole fraction was varied in the calculation of different parts of a global phase diagram. When the points near the joining points between different parts were computed, the step size was set to a rather small value. The step size was then increased to twice of the last value for each subsequent point computed. When the MVC mole fraction was approaching one, the step size was set to a small value to obtain enough points needed to minimize the chances of missing important phenomena. <p>The techniques to set initial guesses for evaluation of different components of a global phase diagram were discussed. The code performance, including the number of iterations for different convergence criteria and the sensitivity of the algorithm were presented. Using the code developed, phase diagrams of type I, type II, type III and type V were generated using representative binary systems from the petroleum processing field. <p>The boundary states between different types of phase behaviour were also explored. It was observed that with the increase of the binary interaction parameters, the phase behaviour of the ethane + ethanol binary system changes from type I to type II to type III while the methane + n-hexane binary system changes from type V to type III. These conclusions matched the results of van Konynenburg and Scott (1980). It was also concluded that with the increase of the binary interaction parameter for a binary system, the system showed a trend to exhibit more liquid-liquid immiscibility.

Flammability limit is a significant safety issue for industrial processes. A certain amount of flammability limit data for pure hydrocarbons are available in the literature, but for industrial applications, there are conditions including different combinations of fuels at standard and non-standard conditions, in which the flammability limit data are scarce and sometimes unavailable. This research is two-fold: (i) Performing experimental measurements to estimate the lower flammability limits and upper flammability limits of binary hydrocarbon mixtures, conducting experimental data numerical analysis to quantitatively characterize the flammability limits of these mixtures with parameters, such as component compositions, flammability properties of pure hydrocarbons, and thermo-kinetic values; (ii) Estimating flammability limits of binary hydrocarbon mixtures through CFT-V modeling prediction (calculated flame temperature at constant volume), which is based on a comprehensive consideration of energy conservation. For the experimental part, thermal detection was used in this experiment. The experimental results indicate that the experimental results fit Le Chatelier’s Law within experimental uncertainty at the lower flammability limit condition. At the upper flammability limit condition, Le Chatelier’s Law roughly fits the saturated hydrocarbon mixture data, while with mixtures that contain one or more unsaturated components, a modification of Le Chatelier’s is preferred to fit the experimental data. The easy and efficient way to modify Le Chatelier’s Law is to power the molar percentage concentrations of hydrocarbon components. For modeling prediction part, the CFT-V modeling is an extended modification of CAFT modeling at constant volume and is significantly related to the reaction vessel configuration. This modeling prediction is consistent with experimental observation and Le Chatelier’s Law at the concentrations of lower flammability limits. When the quenching effect is negligible, this model can be simplified by ignoring heat loss from the reaction vessel to the external surroundings. Specifically, when the total mole changes in chemical reactions can be neglected and the quenching effect is small, CFTV modeling can be simplified to CAFT modeling.