Theses and Dissertations (Mechanical and Aeronautical Engineering)

Permanent URI for this collectionhttp://hdl.handle.net/2263/32359

Browse

Now showing 1 - 20 of 405

Hybrid modelling in a digital-twin context for driveline damage detection and prevention within a mechanised, digitalising mining industry
van Eyk, Luke (University of Pretoria, 2026-02-05)
Modern mining operations rely on large fleets of mechanised mobile equipment to ensure productivity. However, harsh duty cycles make the drivelines on these machines particularly susceptible to failure, establishing them as a primary cause of expensive, unplanned downtime. Against this backdrop, the sector is undergoing a rapid digital transformation, creating an ideal opportunity to leverage digital twins to minimise the impact of these failures. However, industry surveys conducted in this work reveal that the digital twin concept is often misunderstood, confirming the critical need for a standardised definition and approach. To address this ambiguity, this thesis first proposes a digital twin framework tailored for the mining industry. This strategic contribution defines five dimensions of a digital twin, providing a standardised structure for integrating various models to improve decision-making in mining. With this framework established, the work presents two technical contributions that address driveline damage detection and prevention. Both contributions rely on hybrid modelling methods that combine physics-based and data-driven models to provide advanced insights. The first technical contribution addresses reactive damage detection in gearboxes, specifically where historical fault data is scarce. A Simulation-Driven Domain Adaptation (SDDA) methodology was developed to overcome this “few-shot” challenge. In this approach, a generalised physics-based model generates abundant, labelled synthetic fault data, which is used to train a data-driven model. A transfer learning pipeline calibrates and adapts the data-driven model to the real gearbox data domain using limited real-world vibration signals. Validated through controlled numerical investigations and two case studies on real gearboxes, SDDA achieves high fault detection accuracy even with minimal real-world data, successfully classifying faults that purely data-driven and physics-based models fail to detect. The second technical contribution focuses on proactive damage prevention through wheel-slip prediction for skid-steered vehicles. The study finds that standard isotropic friction models are sub-optimal for modelling heavy vehicles’ behaviour on mining terrain, as they miss the directional physics of soil deformation. Consequently, an anisotropic traction model was formulated to explicitly separate the longitudinal and lateral traction coefficients. Integrated with a vehicle state estimator, this method generates real-time available torque envelopes for each wheel or track of a skid-steered vehicle. These envelopes dictate when the wheel is predicted to slip. Field experiments confirmed that this anisotropic approach captures directional traction effects significantly better than the isotropic baseline, enabling timely warnings to prevent damaging wheel slip. Ultimately, this research demonstrates the power of hybrid modelling, combining the structural interpretability of physics with the adaptive calibration of data. By framing these hybrid methods within the proposed digital twin framework, the thesis illustrates a concrete capability to transform raw sensor data into actionable insights for asset failure detection and prevention, paving the way for a more reliable, digitalised mining future.
Influence of vehicle cabin thermal conditions on short-drive cooling demand and passenger comfort
Githinji, Kefitlhile Tshiamo (University of Pretoria, 2025)
This study investigates the thermal interaction between vehicle cabins and their occupants, focusing on personalised cooling strategies that can reduce HVAC energy consumption during short journeys in electric vehicles. The key driver for this study when improving the heating, ventilation, and air conditioning (HVAC) requirements of a vehicle cabin is the relationship between electric vehicles, short journeys, and the need for energy efficiency. The energy expenditure associated with managing the cabin's climatic conditions arises from internal cooling loads and external weather conditions; leading to uncomfortable and hazardous thermal environments for passengers. To optimise the driving range of electric vehicles whilst ensuring passenger comfort, it is essential to study the thermal interaction between the cabin and occupants to implement energy-saving strategies in the vehicle's HVAC system. An experiment was conducted on 31 human subjects, exploring three experimental permutations of different ventilation air flows. One group was subjected to a cooling rate of 0.02 kg/s, the second at 0.03 kg/s, and the third at 0.04 kg/s. Measurements include human subject skin temperature, ambient temperature, subjective responses to changing thermal conditions, and vehicle cabin interior surface temperatures. The study's findings reveal significant variations in individual thermal comfort needs, underscoring the importance of adaptable and personalised climate control systems in vehicles. For example, at an airflow rate of 0.02 kg/s, with 10 subjects, the mean time to discomfort was 36 seconds (±45 seconds standard deviation), with values ranging from a minimum of 1.94 seconds to a maximum of 161 seconds. Additionally, the 25th percentile was 13 seconds, the median (50%) was 13 seconds, and the 75th percentile was 35 seconds. These results reflect the variability of thermal comfort needs even within a single group, emphasizing the importance of flexible climate control. Data analysis showed that the time to notice thermal discomfort varies widely among individuals and depends on the cooling air flow rate, highlighting the subjective nature of thermal comfort. These insights are critical for developing energy-efficient cooling strategies in electric vehicles to enhance occupant comfort and energy efficiency.
Enhancement of flow boiling by the introduction of delta winglet vortex generators
le Roux, Francois Petrus Jacobus (University of Pretoria, 2026-02-01)
Thermal management has been a field of great interest for many years, and is expected to become even more relevant in the coming years, especially with the expansion of computing resources. Phase-changing methods like boiling are identified as a particularly efficient means of removing heat from a surface. This research proposes a simulated investigation into a novel flow boiling enhancement, using delta winglets to produce vortices, to advance the existing cooling methods even more. The simulation is to be conducted in Ansys Fluent. A study of the existing solutions reveal that there is a need for enhanced heat removal methods. Additionally, the study shows that the proposed novel method has not been investigated before, presenting an opportunity for new insight. A literature study is conducted on the fields of boiling and vortex generators, focusing on pool boiling, flow boiling, microchannel flow boiling, longitudinal vortex generators, vortex bursting and enhancements used in flow boiling. From the literature we identify suitable experimental work to use as validation cases for our simulations. Three validation cases are selected.Each validation case is briefly summarised and then simulated. An element of our novel investigation is present in each validation case. The first validation case contains delta winglets, longitudinal vortices and heat transfer. The second validation case contain microchannels and heat transfer. The third validation case contains boiling, via the RPI boiling model, and heat transfer. For each validation case post-processing techniques were developed that we used in the novel investigation. We also showed how our simulation results were within experimental range, and thereby qualified as a validation.For our novel investigation we combined key concepts from the three validation cases into a single simulation, i.e. a delta winglet within a microchannel with flow boiling. We made use of the RPI boiling model to simulate and determine the heat transfer rates. The results of the novel investigation appear promising. Although having a high boiling incipience temperature, the gradient of the boiling curve exceeds that of the validation case.Further study and an accompanying experimental investigation is recommended as future work. A parametric study involving the existing geometric layout and simulation settings would also be able to reveal additional insight into the behaviour of the system.
Efficient construction of spatio-temporal surrogates using gradient enhanced radial basis functions and coordinate transformation methods
Krüger, Stephan (University of Pretoria, 2025)
Surrogate models are increasingly essential for high-fidelity engineering analysis, particularly when response surfaces exhibit anisotropy, strong variable coupling and temporal evolution. Classical radial basis function (RBF) surrogates, while flexible and meshfree, suffer from ill-conditioning and reduced accuracy under such conditions. This thesis develops a comprehensive framework for constructing accurate, stable and data-efficient surrogate models for anisotropic and spatio-temporal applications by integrating gradient-enhanced RBFs (GERBFs) with a curvature-informed coordinate transformation known as the Local Hessian Method (LHM). The research begins with a detailed assessment of Gaussian function value only RBF (FV-RBF) and GE-RBF formulations, demonstrating that gradient information substantially improves interpolation accuracy, especially under sparse sampling. However, the studies also reveal the limitations of classical isotropic kernels in anisotropic domains, motivating the development of transformation-based improvements. The LHM is introduced as a means of reconstructing a representative global Hessian from multiple locally estimated Hessians. The resulting eigenvalue–eigenvector decomposition provides a rotation–scaling transformation that restores isotropy and stabilises the surrogate model. Benchmark studies in two and four dimensions show that LHM-enhanced GE-RBFs match or exceed the performance of Gradient-Enhanced Kriging (GE-Kriging) and the Active Subspace Method (ASM), particularly for coupled anisotropic problems. A further contribution is the extension of the LHM to spatio-temporal domains, where dense temporal sampling creates non-uniform kernel distances and leads to ill-conditioning. The thesis proposes a combined strategy involving centre redistribution, reduced temporal sampling and an improved neighbour-selection scheme for curvature estimation. These developments enable the extraction of meaningful spatio-temporal curvature and yield substantial reductions in prediction error. The final part of the thesis applies the full GE-RBF with LHM framework to a nonlinear finite element analysis of boiler tubes with external erosion defects. Using a four-dimensional spatio-temporal dataset, the LHM-enhanced GE-RBF surrogate achieves an NRMSE of 0.86% and an R2 value of 0.999 on unseen data, outperforming both ASM-based GE-RBF and GE-Kriging at higher training densities. The surrogate is computationally inexpensive and suitable for real-time evaluation, enabling practical integration into inspection workflows for structural integrity assessment. Overall, the thesis establishes best-practice methodologies for constructing GE-RBF surrogates in anisotropic and spatio-temporal settings. The LHM is shown to be an effective and scalable approach for restoring isotropy, improving conditioning and enhancing predictive accuracy. The contributions provide a robust foundation for surrogate-assisted optimisation, structural assessment and future digital-twin applications.
Simulation study for improving heavy payload legged robot locomotion capabilities through the use of a Marxplanar drive arrangement gearbox
Kikanga, Vincent L. (University of Pretoria, 2025-11-11)
This dissertation investigates the feasibility of employing the Marx Planar Drive Arrangement Gearbox (MPDAG) as the primary actuation mechanism for a quadruped robot intended for heavy payload transportation in hazardous environments, such as underground mining. Existing quadruped robots typically exhibit payload-to-weight ratios below unity, rely on complex control systems, and incur high energy costs. To address these limitations, this research develops and evaluates a novel quadruped design that integrates the MPDAG with statically stable gait planning and computationally efficient inverse kinematics. The methodology combines MATLAB/Simulink inverse kinematics with Simscape-based dynamic simulations to assess gait execution, torque and power requirements, and drivetrain stress. Two case studies are performed under a defined 600 kg payload (payload-to-weight ratio of 1.2) conditions. Results demonstrate that the prescribed gait was successfully implemented, with centre of mass trajectories confirming stable locomotion. Dynamic analyses revealed that only one component of MPDAG bears the greatest mechanical demand, with peak torque and power requirements of approximately 6 kNm and 8 kW, respectively. Notably, energy distribution across the legs was uneven, with the right hind leg carrying a disproportionate share of the load. Despite these localized peaks, Lewis bending stress analysis confirmed that gear stresses remained below the material yield threshold (540 MPa vs. 580 MPa), validating the structural reliability of the drivetrain under cyclic loading. The findings establish the MPDAG quadruped as a mechanically viable and energetically efficient design concept, capable of carrying heavy payloads while maintaining static stability. Contributions of the research include the introduction of the MPDAG to quadruped robotics, the implementation of an efficient trajectory generation framework, and the demonstration of drivetrain feasibility using commercially available materials. Limitations include the reliance on simulation, simplified stress modelling, and the rigid body assumption, which restricts load distribution analysis.Future work is recommended to refine gait planning for balanced energy use, incorporate compliant or flexible body structures, apply advanced finite element analysis for detailed stress assessment, and develop a physical prototype for experimental validation. Overall, this study advances the state of quadruped robot design by integrating a parallel gear-driven mechanism with stable gait control, providing a foundation for robust, heavy-duty robots suitable for hazardous industrial applications.
A robust stall and stall-precursor detection method for axial fans
Brocco, Alex (University of Pretoria, 2026-01-01)
Active stall suppression systems possess the ability to prevent the catastrophic failure of main mine fans without the efficiency reductions associated with passive stall suppression systems, such as casing grooves and anti-stall rings. A robust stall detection system is required for the practical implementation of an active stall suppression system in industry. The ability to detect stall precursors would facilitate proactive implementation of the active stall suppression system and prevent the fan from ever entering the mechanically destructive stall condition. To develop a robust stall detection system, an investigation was conducted using experimental data obtained from a 1016 mm-diameter mining ventilation axial fan test-bench, which was developed as a collaborative effort between Air Blow Fans (Pty) Ltd and the Centre for Asset Integrity Management (C-AIM) at the University of Pretoria. The fan test-bench is heavily equipped with a wide variety of sensors, including a zebra-tape shaft encoder and a fast-response pressure probe installed in the rotor casing, in line with 30% of the blade tip chord.Preliminary analysis of the experimental data obtained from the fan test-bench revealed two unique short-length-scale stall precursor types: (i) rotor casing pressure waveform trough magnitude reduction stall precursors and (ii) blade passage pressure dip stall precursors. Both stall precursor types were continuously present during the peak fan pressure operating state. This indicated that they were both linked to rotating instabilities rather than a stall inception mechanism. The trough magnitude reduction stall precursors were only present very close to and at the peak pressure condition, whereas the blade passage pressure dip stall precursors were found at a wide range of operating conditions: from the design point to the peak fan pressure operating state, gradually increasing in strength as the peak fan pressure operating state was approached.The rotor casing pressure waveform trough magnitude reduction stall precursor is postulated to be caused by local blade tip leading-edge flow separation, which results in a reduced pressure differential across the blade tip and a subsequent reduction in the acceleration of the tip leakage flow. The reduced flow acceleration and pressure continuity cause the tip leakage flow pressure to increase (pressure trough magnitude reduction). Experimental evidence suggests that the blade passage pressure dip stall precursors were caused by the low-pressure cores of rotating instability-type radial vortices, which formed in the blade passage just behind the blade tip leading-edge plane and propagated circumferentially to the adjacent blade’s suction-side at part-rotor-speed. These quasi-periodic rotating instability-type radial vortices are postulated to be caused by the interaction between the reversed casing wall flow, the blade tip leakage vortex, and the incoming main passage flow, rather than fluctuating tip leakage vortices. The postulations made relating to the two unique stall precursor types are based on the literature reviewed and the time-averaged, steady-state numerical analysis of the fan test-bench. These stall precursors occurred continuously but stochastically hundreds of revolutions before the onset of rotating stall via spike-type stall inception in the fan test-bench during transient system resistance cases that traversed the fan curve from the stable to the stall operation regions. This provided an opportunity to develop a novel stall and stall-precursor detection method.A novel waveform-difference stall and stall-precursor detection method was developed via the exploitation of advanced signal processing techniques, the fan laws, and the newly obtained understanding of the rotating instability-type stall precursors. The novel waveform-difference method can detect both rotating instability-type and spike-type stall precursors. Additionally, it can detect the presence of rotating stall and surge, regardless of the inception mechanism.The novel waveform-difference method is robust to different steady-state speeds, speed transients, varying levels of noise, and small geometric changes associated with maintenance events and general wear. Additionally, based on its development and the theoretical understanding of how air density changes result in direct magnitude-scaling of the rotor casing pressure waveform’s deterministic component, the method would be robust to changes in ambient air temperature and pressure.The robustness of the novel waveform-difference method theoretically allows it to be calibrated on a single test fan and, thereafter, be applied to all fans in industry that have the same design as the test fan without further calibration, regardless of differences in the operating condition and small geometrical variations. This significantly improves the method’s practicality for industrial implementation. However, additional experiments are required to explicitly prove this one-shot calibration claim. The novel waveform-difference method detected stall 1.582 seconds (39.0 revolutions) and stall precursors 11.564 seconds (285.2 revolutions) before an industry-standard modified Petermann probe detected stall during a transient system resistance case, which traversed the fan curve from the stable to the stall operation regions. An additional benefit of the novel waveform-difference method is its partial robustness to minor fouling since the rotor casing pressure probe utilised has a large diaphragm rather than small pressure taps that get easily blocked by fouling. A limitation of the novel waveform-difference method is its high complexity, subsequent high computational resource demand for real-time implementation, and its case-specific minimum allowable shaft speed limitation (30% of the maximum shaft speed for the fan test-bench utilised). The speed limitation is necessary due to the increase in the relative magnitude of the underlying low-amplitude stochastic pressure fluctuations and low-frequency noise components, compared to the speed-dependent magnitudes of the blade-traversal-induced rotor casing pressure waveform features at low speeds. Lastly, the novel waveform-difference method is tailored specifically for mining ventilation axial fans with thick blades. Therefore, it will be unsuited for other axial turbomachinery types. Additionally, it is best suited for fans that exhibit rotating instabilities and experience spike-type stall inception to facilitate the detection of the associated stall precursors.The algorithm used to implement the novel waveform-difference stall and stall-precursor detection method currently exists in an offline format. Future work will involve developing the waveform-difference method algorithm for real-time implementation. This will involve the implementation of parallel processing and efficient data-handling strategies, as well as tools to facilitate visualisation for remote monitoring. Additionally, future work will include unsteady large-eddy numerical simulations of the fan test-bench to provide more accurate numerical evidence for the postulations made in this work that relate to the two unique rotating instability-type stall precursors.
Sparse parametric data-driven model identification for intermittently and latently forced dynamical systems
Raut, Jay (University of Pretoria, 2025-12-23)
This study explores data-driven, sparse Ordinary Differential Equation (ODE) – based modelling techniques for mechanical machinery, aiming to enhance diagnostic and prognostic capabilities. Effectively using current methods, such as Discrete Fourier Transforms and Kurtosis analysis, requires extensive domain knowledge, whereas interpretable data-driven ODE models offer a more intuitive approach. The Sparse Identification of Nonlinear Dynamics (SINDy) framework is employed to identify sparse parametric ODEs. However, SINDy faces challenges with normalised datasets and non-autonomous systems, such as those with intermittent and latent forces. Two novel heuristics for SINDy are proposed: the first heuristic is a new regression strategy, Sequential Thresholding of the Coefficient of Variation (STCV), which uses a magnitude-free regularisation approach, making it robust against data normalisation. In comparison, STCV represents an advancement over the Ensemble SINDy (E-SINDy) approach. While E-SINDy uses bootstrap aggregating to mitigate overfitting, STCV applies Bayesian Linear Regression for computational efficiency and employs a magnitude-free regularisation approach, enhancing its robustness to data normalisation. The second heuristic allows SINDy to model non-autonomous systems by filtering out significant, temporally sparse errors. This method, named Non-Autonomous SINDy (NAut-SINDy), employs model state prediction error and kinetic energy normalisation, enhancing its application to a broader range of systems. The effectiveness of STCV and NAut-SINDy is showcased through various examples, including simulated Lorenz, Rossler, Van der Poll, and Duffing oscillators and real-world scenarios involving damaged bearings and half-car suspension models. NAut-SINDy's utility particularly emphasises diagnostics, making it possible to identify internal crack shapes in bearings and updating vehicle suspension models online to detect damage. These advancements in SINDy enhance its practicality and accuracy in modelling complex engineering systems.
Initial testing of a parallel-flow microturbine using automotive turbochargers
Humphries, Evan David (University of Pretoria, 2025-12-10)
Micro gas turbines have become increasingly popular for small-scale power generation. The Brayton cycle, on which micro gas turbines are based, requires a heat input – this heat can come from many sources including concentrated solar, petrol and even hydrogen. Single shaft microturbines generally have a large initial cost and a high degree of complexity. The generator is typically attached to the same shaft as the turbine and the compressor, rotating at high speeds of around 80 kRPM and upwards, which means expensive power electronics are usually required. Multi-shaft layouts, including a recently introduced low-temperature turbine (LTT) parallel-flow cycle, simplify some of these issues by using a separate power turbine for the electrical power generation. However, parallel-flow microturbines have not been tested and validated by experiments. The objective of this study was to evaluate the accuracy of analytical and simulated models of parallel-flow microturbines developed from automotive turbochargers by systematically validating them against experimental data. This dissertation is divided into three main sections which include analytical calculations, Flownex® simulations and experimental testing. The analytical approach uses the first principles of thermodynamics, heat transfer and fluid mechanics to predict performance. The simulations were used to initially determine the most promising layouts and were then validated against experimental results. The power turbine in the investigated LTT cycle was limited to a speed of below 40 kRPM, the typical limit of a standard gearbox, therefore the coupling to a generator was made simpler. The parallel-flow LTT combination of the GTX 4088 R and GT 2052 turbochargers was tested experimentally and used to improve and validate Flownex® models of parallel-flow microturbines with piping, gearbox and bearing losses included. The Flownex® model predicted the performance of the experimental setup within 10 % for all performance parameters except heat input, which was within 15.2 %. The Flownex® simulations showed that the combination of the GTX 4088 R and the GT 2052 as the power turbine could have produced 3.97 kW at 1.32 % efficiency while the power turbine is operating at 40 kRPM, but with a gasifier turbine inlet temperature above the maximum allowable temperature of 1050 ˚C. The Flownex® simulations conveyed that the parallel-flow LTT combination of the GTX 4088 R as the gasifier and the GBC 14 as the power turbine could produce 3.26 kW at 1.09 % efficiency, with the power turbine operating at a speed of 40 kRPM and with the gasifier turbine’s inlet temperature below 1050 ˚C. The results showed that this parallel-flow combination can outperform both the single shaft GTX 4088 R and the single shaft GBC 14 when operating at the same speed of 40 kRPM. The validated Flownex® model proved to be a useful tool in predicting the performance of parallel-flow microturbines and can benefit future work on parallel-flow microturbines.
Alternative layouts for coupling a parallel expander to a Brayton cycle configuration developed from commercial turbochargers
Cockcroft, Caitlin Claudia (University of Pretoria, 2025-09)
Microturbines developed from commercial turbochargers may present feasibility for providing electricity to regions with poor power infrastructure. However, turbochargers operate at low pressure ratios where performance is sensitive to the addition of pressure-drop components, which may be required for solar thermal hybridisation and cogeneration. Parallel-flow Brayton cycles have therefore been proposed to reduce the effect of pressure losses on performance when introducing recuperation. This study investigates the feasibility of parallel-flow Brayton cycles using automotive turbochargers, focussing on obtaining maximum thermal efficiency at steady state, while considering various power turbine split-off points. The cycles are investigated by considering turbine inlet air cooling (which can be used for cogeneration via water heating), recuperation, solar hybridisation, and multi-dish setups, with comparisons to single-shaft layouts. Various turbocharger combinations and counter-flow recuperator dimensions are considered, while maintaining turbine inlet temperatures and solar receiver temperatures below the recommended limits. Solar heat inputs are introduced using constant geometry in the form of a 4.8-m-diameter solar dish and an open-cavity tubular receiver, while considering different receiver placements (before the combustor or before the power turbine). Various multi-dish setups are also modelled using two solar receivers. Results show that turbine inlet air cooling extends the pressure ratio range of an unrecuperated parallel-flow cycle and improves performance, particularly in cycles where gasifier turbine inlet temperatures approach manufacturer limits. Of all the recuperated parallel-flow cycles investigated, a recuperated low-temperature turbine (LTT) layout produces the lowest power output but also offers the highest thermal efficiency of 19.2%. For unrecuperated solar cycles, the high-temperature turbine (HTT) cycle with the solar receiver before the combustor provides the best thermal efficiency of 7%. For recuperated solar cycles, the LTT cycle with the solar receiver before the power turbine achieves the highest thermal efficiency of 22%, showing that the LTT cycle offers viability in cycles with many components. In recuperated parallel-flow cycles and recuperated solar parallel-flow cycles, thermal efficiency performance improves under increased combustion pressure losses, from 6% up to 11%, in contrast to the declining performance of single-shaft cycles. At a pressure ratio of 1.8, results show that the recuperated solar parallel-flow LTT configuration can outperform its single-shaft counterpart for combustion pressure losses exceeding 8.7%, with thermal efficiencies increasing from 13.5% to 19.9% with an increase in pressure loss from 6% to 11%. A cycle with a larger solar dish to form a multi-dish setup further enhances thermal efficiency. In the best-performing multi-dish cycle, consisting of the recuperated LTT cycle, the thermal efficiency improves by 69% over the single-dish cycle. Multi-dish parallel-flow configurations have greater solar heat capture through better heat distribution allowing for lower solar receiver surface temperatures. These results demonstrate that parallel-flow Brayton cycles, consisting of automotive turbochargers, can achieve meaningful thermal efficiency improvements over conventional single-shaft configurations, particularly when using optimised split-off points for cogeneration and solar hybridisation. The LTT configuration consistently offers strong performance at low pressure ratios and elevated combustion pressure losses, making it a promising option for small-scale, solar hybridised power generation.
A framework towards quantifying aleatoric and epistemic uncertainties in rolling element bearing condition monitoring experiments
Osetoba, Ayomide D. (University of Pretoria, 2025)
Rolling element bearings (REBs) are critical components in rotating machinery, and their condition monitoring (CM) is essential for predictive maintenance. However, experimental data used for developing and validating CM techniques often contain uncertainties arising from sensor noise, environmental conditions, and test bench disassembly and reassembly. These experimental datasets’ uncertainties, categorised as epistemic and aleatoric, are typically overlooked in developing and validating these CM techniques. This study developed a framework to quantify epistemic and aleatoric uncertainties in REB CM datasets. An experimental methodology was developed to introduce and quantify these uncertainties. The experimental methodology was applied to a bearing test bench at the University of Pretoria. Four cases were designed to quantify epistemic uncertainty introduced by the bearing assemblymodification: (1) baseline configuration (without any modification), (2) bearing preload removal and reinstallation, (3) bearing preload and partial bearing removal (all components except for the outer race), followed by reinstallation of the components, and (4) removal of the preload with the complete bearing whereafter all the components were reinstalled. Aleatoric uncertainty was captured by taking five consecutive measurements per case at fixed intervals under constant speed. The collected signals were further divided into full and segmented windows for comparative analysis and were analysed using the developed envelope analysis-based methodology. Before detailed uncertainty analysis commenced, the influence of key envelope analysis parameters on the results’ interpretation was investigated using a simulated and an experimental signal. These parameters were investigated so that their effect on the experimental uncertainties is minimised. Unlike most studies in the literature, which focus on model-level uncertainties, this work emphasises data-level uncertainties arising from the experimental setup and data collection process. By quantifying and analysing how these uncertainties propagate through feature extraction, the key findings demonstrate that aleatoric uncertainties induce measurable variability between repeatedmeasurements. In contrast, epistemic uncertainty causes significant shifts between cases. This shift is attributable to the changes made to the bearing in the bearing housing, thereby creating out-of-distribution effects on diagnostic features. This shows the importance of accounting for both inherent data variability and experimental setup changes to enhance the trustworthiness and robustness of REBs fault diagnosis technique. This work presents a novel perspective in REB diagnostics, underscoring the importance of data variability and quality. The experimental and analysis framework established here provides the groundwork for integrating experimental dataset uncertainty-awareness into artificial intelligence (AI) based CM techniques (e.g., machine learning).
Optimising maintenance practice through RAMC modelling to extend economic life of ageing railway rolling stock
Daya, Amanuel Degefu (University of Pretoria, 2025)
The economic life of complex systems like railway rolling stock refers to the period over which the asset delivers the most cost-effective service, before the cost of owning, operating, and maintaining it exceeds the cost of replacing it with a new or more efficient asset. It is the duration for which the total life cycle cost per unit of output (e.g., per kilometre or ton-km) is minimised. The economic life of ageing railway rolling stock is a critical determinant of the overall efficiency, reliability, and financial sustainability of rail transportation systems. As rolling stock matures, it typically experiences rising operational failures, declining availability, and escalating maintenance costs – particularly in repairable systems – while renewable components may still be effectively managed through scheduled replacement. These challenges can compromise service quality and profitability. well before the asset reaches its physical end-of-life.This research presents a framework based on a discrete event-based Reliability Availability Maintainability Cost (RAMC) Monte Carlo simulation model to evaluate and extend the economic life of railway rolling stock equipment. The model integrates life cycle costing principles, reliability-adjusted maintenance data, and cost-per-unit-of-output analysis to simulate both "As-Is" and "Best-Case" maintenance strategies. Key system components are categorised into renewable and repairable types to reflect their differing economic implications. Simulation inputs are derived from historical failure, maintenance, and cost records of the rolling stock equipment. The model outputs include the optimal replacement age, total cost of ownership, and equivalent annual cost under varying operational scenarios. The model allows simulating the effects of life extension strategies such as optimising the maintenance regime, changing the operational profile or introducing upgrading of certain systems.The model is validated and use of the model is demonstrated on a real-world case-study, involving Electric Multiple Unit (EMU) rolling stock operated by an African railway company. Results demonstrate that enhanced maintenance strategies can significantly reduce the average annual cost and defer the economically optimal replacement point, thereby extending the asset's economic life without compromising reliability or safety.This study offers a practical, data-driven approach for asset managers, policymakers, and railway operators to make informed investment and maintenance decisions. By identifying the economic tipping point at which continued operation becomes inefficient, this framework supports sustainable life extension of existing rolling stock while minimising financial and operational disruptions.
Steady-state modelling approach for a recuperated hybrid solar-dish Brayton cycle
Phakisi, Tshepo G. (University of Pretoria, 2025-11-18)
A hybrid solar-dish Brayton cycle uses the concentrated power of the sun to preheat air in a solar receiver before it enters the combustion chamber. Such a cycle (recuperated) was considered in the ST-CHP (Solar Turbo Combined Heat and Power) project using an open-cavity tubular solar cavity receiver. Little work has been done to model the receiver for convection heat loss in different locations and environmental conditions (including the effect of dish-shielding of the wind) that may occur in a typical year, owing to the inherent computational cost accompanying such a study. This study therefore investigated the operational nature and parameters of a recuperated solar-dish Brayton cycle including its performance under different climatic conditions in South Africa. A simplified mathematical model was proposed to predict (at a low computational cost) the different modes of heat loss from the cavity receiver, the dish shielding-potential against incident wind and temperature profiles within the cavity. A steady-state semi-analytical model was developed for the receiver using a set of proven correlations from literature and was tested against previous published models and experiments. The solar cavity receiver model was then incorporated into a recuperated Brayton cycle model of the ST-CHP project to include the turbine output power, fuel savings and thermal efficiencies. The HelioClim database containing meteorological data was used as an input to do a case study (using the year 2005) to determine how the dish-receiver system impacts the fuel consumption of the micro-gas turbine. Results showed that the dish plays a significant role in shielding the receiver from high convective heat losses at wind speeds above 2 m/s, at wind directions other than side-on winds. Furthermore, it was found that locations near the coast suffer from higher wind speeds, therefore the dish-receiver system would have a lower thermal efficiency. However, the coastal areas also have a higher atmospheric pressure and lower temperature, which aids the micro-turbine performance. Locations inland, such as Pretoria, have a rich solar resource, and thus the dish-receiver setup is active for longer hours which results in higher fuel savings of up to 15 MWh annually (which translates to 23% fuel savings) when the system operates for 24 hours a day. From the locations considered in this study (for the year 2005), Stellenbosch, Richtersveld, Mariendal and Vanrhynsdorp were identified as the best locations for deployment of the recuperated hybrid solar-dish Brayton cycle with an average overall yearly efficiency of around 20% when operated for 24 hours per day, 21% for 12 hours, and 23% for 6 hours. For comparison, these average overall yearly efficiencies were 10%, 18%, and 37% higher than their no dish-receiver counterparts, respectively. Overall, the model that has been developed can provide significant insights into the ideal locations for deployment of a recuperated hybrid solar-dish Brayton cycle.
Computational analysis of surface tension effects in microchannel flow boiling with self-rewetting fluids
Pienaar, Andre (University of Pretoria, 2024-10)
Continual developments in computing technologies have caused processors and micro-processors to reduce in size while at the same time, operating at higher power densities and increased heat flux demands. Current cooling methods are quickly becoming less effective as they struggle to remove heat at the required rates. Microchannel flow boiling is at the forefront for cooling of high heat flux applications due to its combined convection heat transfer and latent heat transfer mechanisms. However, innovation for flow boiling in microchannels is needed to keep up with the continual developments, particularly because the typical coolant fluids often tend to experience local dry-out regions. There exists a class of fluids, known as self-rewetting fluids (SRF), which possess unique surface tension characteristics that reduce the local dry-out regions. This reduction in local dry-out regions is caused by the fluid motion at the two-phase interface being driven by surface tension gradients as described by the Marangoni effect. Limited information is available to study this phenomenon and therefore any additional information regarding the heat transfer capabilities and fluid dynamics of these fluids is crucial. In this numerical investigation, a 5% v/v 1-butanol-water solution was used as the SRF and studied in a thin horizontal channel at different heat fluxes. This was achieved by conducting two-dimensional (2D) simulations using a domain with a length of 5 mm and a height of 0.3 mm for various applied heat fluxes and an inlet mass flux of 15 kg/m²s. The numerical study investigated the flow of vapour slugs in the channel without the influence of surface wettability by not modelling any contact between vapour slugs and the wall. The results from the SRF were compared to those of water. It was found that the SRF, which has a unique surface tension gradient profile, drew the fluid surrounding the two-phase interface into the hotter region between the heated wall and the vapour slug. The water on the other hand experienced fluid being drawn out of this heated region resulting in thinner liquid films. The difference in the liquid films meant that the hotter fluid near the heated wall was evaporated in the case of water whereas it was trapped in the case of the SRF. Due to the hotter fluid being trapped in the SRF, higher surface temperatures were recorded than in the water case. The higher surface temperatures resulted in the SRF having lower heat transfer coefficients than the water. These results were observed at all the applied heat fluxes. Interestingly, these results are opposite to what was experienced in a similar study with a 0.2wt% heptanol-water mixture as the SRF, where surface wettability was a factor. Here surface wettability refers to the vapour bubble contacting the heated surface, such as during bubble departure, which was not modelled in the current study, instead the bubble was initialised in the fluid stream. The dry-out regions formed were smaller in the case of the SRF than in the water, which caused lower heat transfer coefficients in the water due to the poor heat transfer of vapour. On the other hand, an experimental investigation of the 5% v/v 1-butanol-water mixture as the SRF yielded similar results to the current numerical study in that the SRF experienced lower heat transfer coefficients to the water for a mass flux of G = 15 kg/m2s. The numerical study observed that in the case of slug flow where surface wettability and surface dry-out is not relevant, and where a mass flux of G = 15 kg/m2s is used, the self-rewetting fluid does not provide any clear heat transfer benefits over pure fluids like water. The current work was presented at the following conferences: - The 4th ThermaSMART Workshop titled: “Thermal Management for Net-Zero: Sustainability in Earth and Space Environments”. The workshop was hosted by the University of Pretoria from the 10th – 11th August 2023 at Lagoon Beach Conference Centre in Cape Town, South Africa. The work was presented by Mr André Pienaar. - The 76th Annual Meeting of the Division of Fluid Dynamics by the American Physical Society (APS/DFD). The APS/DFD meeting was held at the Washington Convention Center in Washington DC, United States of America from the 19th – 21st November 2023. The work was presented by Mr André Pienaar.
Computational study of cryogenic jet impingement boiling on concave and singular needle surfaces
Somerville, Daiman Athony Hugh (University of Pretoria, 2025-02)
Jet impingement boiling is a highly effective method of surface cooling and is particularly suited to high heat-flux applications such as microprocessor cooling and cryogenic probe cooling. In the present study, we computationally explore the effect that the geometric parameters of surface curvature, needle height and impingement height have on the boiling curve. The multiphase Eulerian model coupled with the {Rensselaer Polytechnic Institute (RPI) wall boiling model is employed in tandem with an axisymmetric steady-state assumption. {The RPI wall boiling model decomposes the wall heat flux into three components: quenching, convective and evaporative. Each of these components comprise of multiple closing models.} A focus is placed on selecting RPI closing models that are suitable for cryogenic fluids. Moreover, a new empirical correlation is proposed for the non-dimensional area of influence {(A_b) which determines the ratio of quenching to convective heat flux.} Based on results of a Monte Carlo experiment aimed at mimicking the distribution of nucleation sites and their overlapping area of influence, { this model is given as: A_b=1-e^{-1.5\beta^{1.2}}. The bubble waiting time coefficient (C_{wt}) can be described as the quenching heat flux pre-multiplicative coefficient and is responsible for adjusting the quenching heat flux independent of the evaporative and convective components. The range for which the bubble waiting time coefficient (C_{wt}) can vary is likewise computed using the results of the Monte Carlo method with the inclusion of the reduction of area of influence due to bubble growth (1
Experimental investigation of film-cooling hole performance
Dlamini, Zimase (University of Pretoria, 2020-06)
Film cooling has, over the years, allowed for the operation of modern gas turbines at temperatures far exceeding the limits of the material properties of the turbine components. This has resulted in increased power output and efficiency of the gas turbines. But over 40+ years of research has not culminated in the goal of achieving ideal cooling films, such as from two-dimensional (2D) continuous slots. This study employed a curvature in the forward diffuser section of the film cooling hole; these holes are referred to as cases 1 to 4 in this study. This was expected to improve the performance of the hole. The performance parameters investigated and reported were the discharge coefficient of the holes, the flowfield downstream of the hole exit trailing edge, the temperature field downstream of the hole exit trailing edge and the effectiveness. The effects of pressure ratio, mainstream crossflow, compound angle, hole geometry, manufacturing method, 3D print build orientation, and inclination angle, on the discharge coefficient were investigated. The effects of blowing ratio, hole geometry, compound angle, turbulence intensity and downstream distance from hole exit trailing edge, on the flowfield, temperature field and effectiveness were also investigated. The hole geometries had a diameter of 8 mm and length to diameter ratio equals to 7.5. The compound angle was varied between zero (0) to sixty (60) degrees. The inclination angles of the holes were either thirty (30) and forty (40) degrees. The effect of the compound angle, manufacturing method and 3D print build orientation was found to be negligible for the discharge coefficient. But the above parameters had a significant effect on the adiabatic film cooling effectiveness. Cases 1 to 4 holes showed higher discharge coefficient values as compared to the cylindrical and the laidback fan-shaped holes. This was a result of the development of the flow inside the hole and the resulting exit coolant jet velocity profile and its interaction with the mainstream crossflow. From the flow structure and temperature field measurements it was determined that employing the curvature and the lateral expansion of the cases 1 to 4 holes decreases the height and trajectory of the jet on exit. The decreased height is due to the decreased vertical momentum content of the coolant jet. The decreased trajectory positions the longitudinal vortices closer to the wall which results in better lateral spread of the coolant. From the effectiveness measurements it was found that increasing the compound angle decreases the lateral averaged effectiveness. And a decrease in the lateral averaged effectiveness was observed as the blowing ratio was increased. The case 2 hole geometry resulted in low jet height when in the mainstream, which means that it was closer to the surface that requires cooling. It also resulted in a relatively good lateral spread of the coolant on the surface. And it resulted in the highest laterally averaged effectiveness at most of the compound angles and blowing ratios tested.
Experimental investigation of magnetic hybrid nanofluids for enhancing heat transfer in forced convection internal flow through transition and turbulent regimes
Adogbeji, Victor Omoefe (University of Pretoria, 2024-12)
Efficient heat transfer is critical for optimizing the performance and safety of industrial and engineering systems. While nanofluids have demonstrated Outstanding heat transfer efficiency juxtaposed to deionized water (DIW), the exploration of Magnetic Hybrid Nanofluids (MHNFs) in forced convection heat transfer within transition and turbulent flow regimes remains limited. This study investigates the thermal and hydrodynamic characteristics of MHNFs, specifically Fe_3 O_4/TiO_2, Fe_3 O_4/MgO, and Fe_3 O_4/ZnO, flowing through a heated pipe. The research examines their performance across laminar, transition, and turbulent flow regimes, with suspension concentrations between 0.00625% and 0.3%. The study was conducted in four phases, addressing critical aspects of MHNFs stability, thermophysical properties, and heat transfer dynamics under varied conditions. In the first phase, the effects of hybridization mixing ratio (HMR), nanoparticle size, and temperature on the stability and thermophysical properties of Fe_3 O_4/TiO_2-DIW, Fe_3 O_4/MgO-DIW, and Fe_3 O_4/ZnO-DIW were evaluated. Results showed that Fe_3 O_4/ZnO-DIW at an 80:20 HMR demonstrated the highest thermal conductivity enhancement (31.28%) and lowest viscosity at 50°C, achieving the best thermal conductivity-viscosity balance. Fe_3 O_4/TiO_2 (18 nm)-DIW exhibited the highest electrical conductivity (4.23 mS/cm) at 50°C. Temperature emerged as the most influential factor on thermal conductivity, emphasizing the potential of MHNFs for advanced cooling applications, such as in proton exchange membrane (PEM) fuel cells. The second phase analysed the heat transfer capabilities of Fe_3 O_4/TiO_2 fluids across Reynolds numbers and volume fractions. Significant improvement in the convective heat transfer coefficient (CHT) were observed, particularly at lower concentrations; at 0.3 vol.% is 11.42% , at 0.2 vol.%, is 14.03% , at 0.1 vol.% is 18.04%, at 0.05 vol.% is 19.98%, at 0.025 vol.% is 22.91%, peaking at at 0.0125 vol.% is 26.33%, and at 0.00625 vol.% is 24.30%. at 0.3 vol.% (21% at Reynolds number 5019) the pressure drops were highest and progressively reduced with lower concentrations: 13.10% at 0.2 vol.%, 11.94% at 0.1 vol.%, 9.82% at 0.05 vol.%, and minimal at 0.0125% and 0.00625 vol.%. The Total Efficiency Index (TEI) was maximized at 0.0125 vol.%, indicating the optimal balance between heat transfer enhancement and minimal hydraulic resistance. The third phase focused on Fe_3 O_4/MgO MHNFs, revealing unique thermal transport behavior in the transition region. Results indicated delayed transition at higher Re juxtaposed to DIW. The thermal transport enhancements were observed across volume fractions, with increases of 26% at 0.3 vol.%, at 0.2 vol.% is 25.8%, at 0.1 vol.% is 25.7%, at 0.05 vol.% is 17.9%, at 0.025 vol.% is 25.6%, at 0.0125 vol.% is 31.6%, and at 0.00625 vol.% is 30.2%. Optimal performance was achieved at 0.0125% and 0.00625 vol.%. However, higher concentrations resulted in increased pressure drops, demonstrating the intricate relationship between fluid dynamics and thermophysical characteristics. In the final phase, the effects of magnetic field strength and waveforms parameters were investigated for Fe_3 O_4/TiO_2 nanofluids. Magnetic waveforms sine, square, and triangular enhanced heat transfer, with increases of 27.87%, 28.21%, and 26.74%, respectively, at 0.0125 vol.%. Magnetic field frequency (40 Hz to 1000 Hz) and voltage (2V to 12V) demonstrated a direct correlation with thermal performance, with optimal results achieved at 60 Hz and 4V. These findings advance the understanding of MHNFs in forced convection heat transfer, offering practical insights for thermal management in power generation, HVAC systems, and chemical processing. The demonstrated ability of MHNFs to optimize heat transfer efficiency while minimizing pressure losses at lower concentrations presents a promising avenue for improving energy efficiency in heat exchangers and thermal systems.
A methodology for risk-based management of the remaining life of structures in vertical mine shafts through integrated structural inspection and maintenance
Ngcobo, Glory Nomvula (University of Pretoria, 2024)
Mine shaft structures play a crucial role in mining operations, and their structural integrity is essential for ensuring operational safety. However, these structures deteriorate over time because of various factors such as corrosion, cracking, structural damage and structural ageing. The structural deterioration impacts on production, safety, and the environment. Estimating their remaining life is critical for preventive maintenance planning and ensuring continued operation. Corrosion significantly affects steel structures, leading to accelerated deterioration and potential catastrophic failures. Regular Structural Inspection and Maintenance Management (SIMM) programmes are essential to detect and address wear, preventing failures, and extending shaft lifespan. Challenges persist in managing corrective measures for mine shaft maintenance, hampering data reliability and informed decision-making. The disconnect between structural inspection results and maintenance systems leads to unaddressed defects and inadequate tracking of critical maintenance in Computerised Maintenance Management Systems (CMMS). This gap hampers informed decisionmaking regarding life cycle cost (LCC) and life of mine (LoM) estimation. This dissertation aims to bridge the gap by integrating structural inspection and maintenance data with CMMS, proposing the Integrated Structural Inspection and Maintenance Management System (iSIMM) and prognostic and risk/economics-based maintenance decision-making model for improved structural maintenance in the mining industry. The research journey starts by exploring mine shaft intricacies and the SIMM, followed by an extensive literature review on structural health and monitoring, covering corrosion, ageing, maintenance, and regulatory requirements. It progresses to developing an iSIMM and a maintenance decision-making model. Validation through a case study at Harmony Gold Mine confirms the practical application of these models in real mining environments. The research findings underscore the potential extension of mine shaft structures’ lifespan by timely replacement of degraded parts, ensuring continuous operations without significant disruptions. Investors have the opportunity to assess risks and returns for possible mine life extension. The implementation of an integrated inspection system enhances visibility into required maintenance for the mine shaft structures, facilitating effective monitoring of maintenance progress and structural condition changes affected by proper upkeep. Proactive monitoring and management practices contribute to 3 reducing LCCs by employing risk-based inspection and condition-based maintenance strategies for optimal maintenance. Moreover, the prognostic and risk/economics-based maintenance decision-making model aids in estimating mine life and supports resource planning. Predictive modelling plays a vital role in determining replacement costs, contributing to future budget planning efforts. The credibility and dependability of the findings are based on the analysis of reports from third-party structural engineering inspections conducted over the past decade. These findings have been verified through physical inspections, in-depth discussions with both mine engineers and contracted structural engineers, and consultations with the engineering teams responsible for performing regular shaft examinations.
Investigating aerodynamic factors in wind-driven grey-headed albatross crash-landings
Schoombie, Janine (University of Pretoria, 2024-10)
A significant portion (ca. 10%) of the global population of Grey-headed albatrosses (Thalassarche chrysostoma) breed on sub-Antarctic Marion Island. It has been observed that a large number of adult grey-headed albatrosses (GHAs) crash into the valley below an important inland breeding site, which proves fatal in most cases. The possible effect of the prevailing wind on the frequency and spatial distribution of crashes was unknown, prompting an investigation into the aerodynamic limitations of GHAs and their interaction with fine-scale wind patterns around this inland breeding site. The purpose of this study is to investigate the frequency of adult GHA crashes and identify any spatial patterns in these incidents. These data would then be compared with both measured and simulated wind vectors to determine if wind plays a significant role. Additionally, the study aimed to examine the aerodynamic limitations of the birds, specifically relating to lift generation, to ascertain which wind conditions are most likely to result in crashes. Field observations were carried out on Marion Island to study the way the GHAs interact with their environment. Flight observations were conducted to directly witness GHA crash-landings and link them to specific prevailing wind conditions, which were measured near the study area. The GHAs nest in dense colonies along an east-facing ridge called Grey-headed Albatross Ridge (maximum height of ca. 200 m a.g.l.). Carcass surveys were conducted in a 1 km2 area below the Ridge, spanning the length of this sub-colony (ca. 4000 breeding pairs) for four breeding seasons (2017-2020). Measured wind data were supplemented by computational fluid dynamics (CFD) simulations of wind vectors over Marion Island (Goddard et al., 2022) that were compared to the spatial distribution of carcasses. Additionally, an aerodynamic investigation of a GHA in flight was conducted using CFD. No geometry for a GHA wing (or body) was available at the initiation of the study except for a few studies noting the wingspan, aspect ratio, weight and planform of an adult GHA. Thus, a GHA geometry was created based on a dried GHA wing and photographs of GHAs in flight. The CFD simulations were conducted using Simcenter Star-CCM+ at a range of speeds, angles of attack and sideslip angles. Observations of albatrosses in flight indicate that most birds are killed when attempting to leave the colony, specifically when flying low above ground in strong wind conditions. During the prevailing westerly winds, the GHA chicks were sheltered, while adults flying low over the valley experienced highly variable wind vectors. In winds approaching from the north, east or south, GHAs were able to take off with a headwind to gain altitude after take-off. The spatial distribution of carcasses also ii showed a clear hotspot in an area of highly variable flow in the lee of the Ridge during the prevailing wind conditions. The CFD investigation showed that a GHA flying with fully extended wings can generate lift up to 9 times its weight at airspeeds above 35 m.s-1 and positive angle of attack (updrafts). A GHA in this gliding posture can generate sufficient lift at a minimum speed of 13 m.s-1 and angles of attack above 10°. For a given speed and angle of attack, the lift decreases with increasing sideslip angle. At high sideslip angles and negative angle of attack, negative lift is generated, resulting in downward acceleration. At high speeds, this downforce can be equal to the bird’s weight, most likely causing fatal injuries if forced to crash. A combination of high cross-winds and low-to-moderate downdrafts is most common in the valley during the dominant westerly winds. Most cases of crash-landings of low-flying GHAs occur during these conditions. Additionally, most GHA carcasses were located in areas where westerly winds create variable airflow in the Valley, leading to low or negative lift generation that increases crash risk, especially during departure flights towards the ocean. GHAs taking off in Westerly winds tend to lose altitude quickly, exposing them to cross-winds and downdrafts for much of their 2 km journey, raising the likelihood of crash-landings. By contrast, northerly, southerly, or strong headwinds allow GHAs to gain altitude faster, reducing the risk of crash-landing. It is proposed that the primary causes of crash-landings are a combination of variable wind vectors and low altitude: at low heights, GHAs lack sufficient time to recover from adverse gusts, even with flapping. Anthropogenic climate change affects wind patterns in the Southern Ocean, impacting GHA foraging and nest accessibility on Marion Island, where increased northerly winds create challenges for birds reaching inland nests. Northerly winds pose both risks and benefits, as moderate northerlies can improve take-off conditions. However, increased frequency and intensity of strong northerly winds could hinder feeding and chick-rearing, threatening the breeding success of this already endangered species. While continued surveys are required to monitor the effect of changes in wind conditions over the long term, the current level of wind-related mortality in GHAs is substantial for a long-lived species with a low natural mortality rate. This is the first study to document persistent wind-driven, land-based mortalities in albatrosses. It presents a unique insight into the way seabirds are affected by wind, not only in the open ocean, but also on land around breeding sites.
Effects of noise and measurement error on load reconstruction of dynamic loads in the frequency domain
Kruger, Johannes Adriaan; Kruger, Johan A. (Adriaan) (University of Pretoria, 2024-08)
Load reconstruction involves solving the inverse problem by using measured structural responses to determine the forces acting on a structure. The inverse problem is almost always ill-conditioned making solving the inverse problem more difficult compared to the forward problem. Response measurements will always contain a level of noise even under the best of conditions. Various methods of performing load reconstruction have been developed in the time and frequency domains. A common problem encountered by all load reconstruction methods are their sensitivity to noise, because of the ill-conditioned nature of the inverse problem. The frequency domain methods have the benefit of limiting the band in which load reconstruction is performed which can eliminate some of the effects of noise. However, frequency domain methods remain sensitive to noise and in certain practical applications noise is present in the same frequency band in which the loads are to be reconstructed. From the literature reviewed, frequency domain methods are less sensitive to noise and have the benefit that they can be performed with very little prior knowledge of the system. The aim of this research was to establish the sensitivity of several commonly used load reconstruction methods used in engineering applications. In order to accomplish this, literature was reviewed to gain insights into the field of load reconstruction. Mathematical models were created to test the sensitivity of selected load reconstruction methods. These models consisted of a lumped mass analytical model, a rigid body simulation model and an elastic beam model which was replicated experimentally. The feasibility of using finite element modelling in conjunction with experimental data to perform load reconstruction was also briefly evaluated. It was found that systematic errors posed the greatest risk as in many of the test cases, no clear indications of any error were noticeable. Meaning that there is a great risk of drawing inaccurate conclusions when systematic errors are present in measured data. In terms of stochastic noise, it was found that the Tikhonov methods were the best performing methods. These methods were able to maintain accuracy up to high noise levels. It was also found that it is feasible to make use of finite element analysis in conjunction with experimental data to perform load reconstruction. However, the accuracy of the finite element modelling plays a major role in the accuracy of the load reconstruction.
Testing and development of a solar-dish cavity receiver for the melting of zinc metal
Bezuidenhout, Pieter J.A. (University of Pretoria, 2024-08)
Concentrating solar power technologies can be applied to reduce the cost and carbon footprint of zinc melting processes. This study aims to improve the knowledge related to small-scale solar melting using a dish concentrator. This technology can be applied to zinc production as well as a range of small-scale applications, such as casting, recycling, galvanisation, and thermal storage. An experimental and analytical analysis of a rotating cylindrical cavity receiver for the indirect melting of zinc metal using concentrated solar power is presented. A multi-facet parabolic dish with an incident area of 2.85 m² was considered together with a rotating cylindrical cavity receiver. The receiver had an aperture diameter of 0.2 m and the capacity for housing 17 kg of zinc. Five experimental test runs were executed, during which up to 73.5 % of the zinc inventory could be tapped from the receiver in its molten state, and average thermal efficiencies of up to 42 % were achieved. A predictive analytical model considering wind speed, wind direction, and direct normal irradiance was developed and validated against experimental data. A heat transfer efficiency factor was experimentally determined to account for voids in the zinc feedstock. The model was used to predict that approximately 41 kg of molten zinc could be tapped from the experimental setup throughout a typical day with a peak direct normal irradiance of about 900 W/m² and an average wind speed below 2 m/s. A case study highlighted that energy savings of 0.6 kWh are achievable per kilogram of zinc processed by concentrated solar power rather than the conventional induction furnace.

Browse

Recent Submissions