Includes abstract. / Includes bibliographical references (leaves 115-117). / Smooth Particle Hydrodynamics (SPH) is one of the simplest meshless methods currently in use. The method has seen significant development and has been the germination point for many other meshless methods. The development of new meshless methods regularly uses standard SPH as a starting point, while trying to improve on issues related to consistency and stability. Despite these perceived flaws it is favoured by many researchers because of its simple structure and the ease with which it can be implemented.

Includes abstract. / Includes bibliographical references. / In this paper, a multi level contextual classification approach of the City of Cape Town, South Africa is presented. The methodology developed to identify the different objects using the multi level contextual technique comprised three important phases.

This dissertation reports on a series of experiments and numerical simulations that were carried out to investigate the response of quadrangular plates to buried encased charges with a view of understanding landmine effects on simple structures. Different loading scenarios were carried out for comparison. In total, four loading scenarios were investigated; namely bare and encased charges detonated in air, and bare and encased charges detonated under sand. A vertical pendulum was used to measure the impulse imparted onto the target plates, and dry construction sand was used to bury the charges. The geometric scaling of the target plates and PE4 explosive charge was based on the Casspir APC and TM-57 Anti-tank mine respectively. The experiments were carried out under varying conditions, such as different standoff distances (50-90 mm) and depths of burial (0-10 mm). In general, the impulse and midpoint deflection decreased with an increase in total distance for all loading scenarios. Burn diameters were observed on plates loaded in air, with dissimilar scorch areas observed on plates subjected to buried charges. Scorched areas on plates subjected to encased charges indicated a focussing effect of the explosive products provided by the charge casing. Plates subjected to encased charges detonated in air typically resulted in ’capping’ in the central area, accompanied by significant shrapnel damage. In general, presence of a charge casing in buried charge tests resulted in more damage to the target plate. Larger impulses and midpoint deflections were measured for charges detonated under sand compared to bare charges detonated in air. Encased charges detonated under sand resulted in decreased impulse imparted onto the target plate, accompanied by an increased plate midpoint deflection when compared to bare charges detonated under sand. The presence of sand in encased charge tests tended to mitigate the shrapnel damage caused by the charge casing. ANSYS AUTODYN was used to perform numerical simulations on three variable standoff distance test series, one with no sand and a bare charge, one with a constant depth of burial and a bare charge, and the last with a constant depth of burial and an encased charge. These simulations mirrored the experimental test series. The simulations consisted of two-dimensional axisymmetric models, with lagrangian meshes representing the target plate and explosive casing, and an eulerian mesh used to model the behaviour of the PE4, sand and air. The blast was simulated in two phases, namely detonation of the explosive and loading of the structure. The casing mesh was only included in the detonation phase. Two separate models were used to simulate the impulse and the plate behaviour. The impulse model used a reflective boundary to represent the plate and pressure histories on the reflective boundary were used to calculate the impulse. The plate loading model included a lagrangian mesh to represent the plate and simulate its deformation. The plate loading model used and additional unloading phase to allow plate oscillations to subside. The numerical model was validated using the impulse, plate midpoint deflection and plate profile measured during the experiments. The numerical model showed good correlation with the results of the experiments in terms of midpoint deflections and impulse trends. The model provided insights into the experiments, such as how the gas products expanded after detonation and their interaction with the target plate.

Forest fires are a pervasive and serious problem. Besides loss of life and extensive environmental damage, fires also result in substantial economic losses, not to mention property damage, injuries, displacements and hardships experienced by the affected citizens. This project proposes a low-cost intelligent hyperspectral 3U CubeSat for the production of fire risk and burnt area maps. It applies Machine Learning algorithms to autonomously process images and obtain final data products on-board the satellite for direct transmission to users on the ground. Used in combination with other services such as EFFIS or AFIS, the system could considerably reduce the extent and consequences of forest fires.

For many mobile robotic systems, navigating an environment is a crucial step in autonomy and Visual Simultaneous Localisation and Mapping (vSLAM) has seen increased effective usage in this capacity. However, vSLAM is strongly dependent on the context in which it is applied, often using heuristic and special cases to provide efficiency and robustness. It is thus crucial to identify the important parameters and factors regarding a particular context as this heavily influences the necessary algorithms, processes, and hardware required for the best results. In this body of work, a generic front-end stereo vSLAM pipeline is tested in the context of a small-scale outdoor wheeled robot that occupies less than 1m3 of volume. The scale of the vehicle constrained the available processing power, Field Of View (FOV), actuation systems, and image distortions present. A dataset was collected with a custom platform that consisted of a Point Grey Bumblebee (Discontinued) stereo camera and Nvidia Jetson TK1 processor. A stereo front-end feature tracking framework was described and evaluated both in simulation and experimentally where appropriate. It was found that scale adversely affected lighting conditions, FOV, baseline, and processing power available, all crucial factors to improve upon. The stereo constraint was effective for robustness criteria, but ineffective in terms of processing power and metric reconstruction. An overall absolute odometer error of 0.25-3m was produced on the dataset but was unable to run in real-time.

In this thesis, a proof of concept receiver system operating from 14.5 to 20 GHz for the MeerKAT Radio Telescope is presented. MeerKAT is a 64 element telescope antenna array consisting of offset-fed Gregorian reflector antennas with a 13.5 m main reflector and 3.8 m sub-reflector. Currently, the MeerKAT is planned to operate up to 14.5 GHz. However, the reflector surface accuracy of 0.6 mm RMS achieved for the MeerKAT potentially allows it to operate at much higher frequencies. The system design consists of a feed horn antenna and front-end down conversion receiver ready for integration with back-end digital signal processing. The antenna design was carried out using electromagnetic simulation software and system level simulation software was used for the front-end receiver. A single polarization wide-axially corrugated horn with low side-lobes and cross-polarization has been designed for the proof of concept with a predicted aperture efficiency of 60% including surface accuracy loss when illuminating the MeerKAT reflector. The measured results for the antenna show a return loss better than 15 dB in the operational band and boresight gain of 12 dB. The measured E- and H-plane cross-polarization for the antenna is lower than -40 dB. The measured edge taper at the halfsubtended angle of the sub-reflector is between -11.8 dB and -13.2 dB. The front-end receiver was designed to use a single down-conversion stage to a 4.5 GHz IF with an instantaneous bandwidth of 2.5 GHz to be bandpass sampled at 6 Giga-samples per second (GSPS). The receiver was designed using off-the-shelf connectorized modules and custom designed microstrip filters for image rejection and anti-aliasing. Laboratory measurements of the receiver show a maximum gain of 76 dB, 40 dB image rejection and 27 dB spurious free dynamic range (SFDR). The simulated noise figure of the system using the measured noise figure of the LNA is 1.74 dB. The measured gain flatness of the receiver is ±7 dB due to poor performance of one of the amplifier modules used in the system.

Includes bibliographical references. / The research focus of this thesis is lifter height effect on different charge characteristics. Thecharge characteristics identified were shoulder and toe angle, charge size and toe height. Anattempt was made to develop a model for the velocity profile incorporating lifter height.Therefore the objectives of the study are to: •Determine the effect of lifter height on the velocity profile using the particle tracking data. •Determine the effect the lifter height on the charge shoulder and toe. •Develop a velocity profile model including lifter height using granular flow theory and to compare the model to experimental data. The motivation for the study is that lifters are one of the most important design variables in a mill. Without lifters the mill’s energy efficiency would decrease. In Meaders & MacPherson, (1964) the effect of lifters on energy was quantified to be between 20% and 30%. The lifters control the height and angle of departure from the charge at the mill shell and therefore control the impact area and magnitude. The area and magnitude of the charge impact will affect the grind of the mill and energy utilization in the mill. The thesis was also aimed at generating data that can be used to model the velocity profile that can be incorporated in power models. Most of the power models do not account for the effect of lifters. The experiments involved collecting data from three dimensional particle tracking of selected particles in the charge. using the PEPT system. The PEPT experiments were conducted at IThemba Labs in Cape Town South Africa using a 300 mm x 285 mm experimental mill. The charge used for the experiments were glass beads with an approximate specific gravity (SG) of 2.7. The PEPT system operates by tracking the x, y and z coordinates with respect to time of an irradiated particle (tracer). The experiments were run under different conditions to evaluate the effect of the lifter height. The experiments were operated by varying mill speed (55%, 70% and 85%), mill fill (20%, 30% and 40%) and lifter heights (1.5 mm, 3 mm, 6 mm and 10 mm).

Includes abstract. / Computational modelling is fast becoming the chosen way to predict novel structures. These structures need to be validated in order to gain credibility and generate more confidence in computational predictions. This investigation of the equiatomic region of the Ag-Pt system shows that computational modelling can be used successfully as a precursor to experimental investigations. Various techniques including electron microscopy, hardness and Differential Scanning Calorimetry were used to investigate different properties of the alloy. These techniques have shown that an ordered phase may exist in the Ag-Pt system. This ordered phase has been shown to have an increased hardness and has produced extra reflections in electron diffraction patterns. Scanning Electron Microscope equipped with a Backscattered Electron detector has shown that a third phase is present in the alloy and the composition is close to 50:50; within experimental error. The alloys showed considerable inhomogeneity and it was not homogenised prior to or post cold rolling. This could be a reason for the third phase not reaching an equilibrium state after prolonged heat treatments. The final structure might be the L11 structure but the full transformation is slow and further investigation is required. It is recommended that future research be carried out taking into account the recommendations provided in Chapter 7.

Includes abstract. / Includes bibliographical references (p. 84-87). / Fischer Tropsch Synthesis (FTS) is the conversion of syngas (CO and H2) to cleaner liquid transportation fuels. The modelling of such a supercritical, highly non-ideal, multi-component system requires the detailed knowledge of the reaction mechanism, reaction kinetics, phase-equilibrium and reactor technology. The objectives of this work were to: develop a complete Fischer Tropsch model, predict the hydrocarbon product distribution, determine what effect Vapour-Liquid Equilibrium (VLE) has on the product distribution, selectivity and kinetics, and determine whether the deviations from the 'ideal' Anderson-Schulz Flory (ASF) distribution can be attributed to VLE.

Summary in English. / Includes bibliographical references. / The purpose of this research is to investigate, review and analyse the use of cloud resources for the enhancement of mobile computing. Mobile cloud computing refers to a distributed computing relationship between a resource-constrained mobile device and a remote high-capacity cloud resource. Investigation of prevailing trends has shown that this will be a key technology in the development of future mobile computing systems. This research presents a theoretical analysis framework for mobile cloud computing. This analysis framework is a structured consolidation of the salient considerations identified in recent scientific literature and commercial endeavours. The use of this framework in the analysis of various mobile application domains has elucidated several significant benefits of mobile cloud computing including increases in system performance and efficiency. Based on recent scientific literature and commercial endeavours, various implementation approaches for mobile cloud computing have been identified, categorized and analysed according to their architectural characteristics. This has resulted in a set of advantages and disadvantages for each category of system architecture. Overall, through the development and application of the new analysis framework, this work provides a consolidated review and structured critical analysis of the current research and developments in the field of mobile cloud computing.