7490 research outputs found
Sort by
Optimisation techniques to improve the drive efficiency of a mobile ventilator platform
Get PDFCOVID-19 pandemic has underscored the indispensable role of mechanical ventilators in providing critical respiratory support to patients. This research has focused on the imperative task of optimising the articulation system of the ventilator, designed to enhance its efficiency, with particular emphasis on improving the volume displacement from the ambu bag. Several optimisation methods were explored, including generative design (GD), Design of Experiments (DOE), Shape optimisation, and topology optimisation. The study also highlights the significance of additive manufacturing and material testing in developing ventilator components. The study delves into the intricate development and fine-tuning of the ventilator setup, emphasising its pivotal role in delivering life-sustaining respiratory aid. The ventilator’s core mechanisms, featuring a two-pusher arm system powered by a servo motor, was engineered intricately to apply precise pressure on the ambu bag. The research underscores the importance of optimising both the pusher arm and pressure plates to improve air displacement within the system. A significant challenge addressed in this research was the excessive strain on the servo motor owing to the demands of the articulation system. The research employed strategies such as shape optimisation and topology optimisation to reduce the stress on the articulation system while increasing the air displacement and thus reducing the pusher arm displacement on the ventilator. The research methodology included stages such as setting performance benchmarks, calibration, and verification to ensure precision and reliability; shape optimisation for maximum efficiency; and topology optimisation for superior structural performance and reduced weight. These interconnected stages were instrumental in the comprehensive development and enhancement of the ventilator system, ensuring its effectiveness and dependability in delivering lifesaving respiratory support.This research extensively examined sensor reliability and performance through verification tests and calibrations, highlighting the precision of the servo motor and the suitability of the 5-Amps current sensor for monitoring servo motor current without additional calibration. Optimisation efforts aimed to enhance the ventilators performance by relocating the pusher arm to the bag’s centre, resulting in improved volume displacement efficiency by 7.78 % and a 25.35 % reduction in current consumption. Shape optimisation, especially with curvature-based pressure plates, increased volume displacement by 84.47 % reaching an optimal configuration outputting 1475.73 ml of volume per compression. Understanding the forces through strain gauges and FEA facilitated topology optimisation, the MAXSTIFFDS15 configuration demonstrated promising results by reducing component weight and achieving significant energy savings of 45.04 %, potentially reducing long-term costs.Thesis (MEng) -- Faculty of Engineering, the Built Environment, and Technology, School of Engineering, 202
Comparing stable isotope ratios and metal concentrations between components of the benthic food web: a case study of the Swartkops Estuary South Africa
No full textEstuarine systems are highly productive ecosystems; however, they are subjected to high anthropogenic pressure such as metal contamination and: increased nutrient loads. The contamination sources of metals and nutrients in urban estuaries are derived: from industrial waste. agricultural and urban runoff that flows into estuaries. An example of such a system is the Swartkops Estuary. industry and three wastewater treatment plants within the Swartkops River catchment are major sources of metal. and nutrient pollution, respectively. The metals accumulate in the environment, are biomagnified up the food web, and transferred from one trophic level to another. At lethal concentrations, metals pose a threat to organisms using the estuary by affecting their physiological and biochemical processes. Stable Isotope analysis has proven to be an effective tool for investigating, trophic linkages in the food chain from a variety of environments. By assessing both metals and stable _isotopes in the. estuary it can provide a more robust understanding of the pathway metals accumulate, biomagnified, and transfer from the environment through the estuarine food web.Thesis (MSc) -- Faculty of Science, School of Environmental Sciences, 202
Seagrasses in Mozambique: distribution, people’s perceptions and restoration best practice
Get PDFThis thesis investigated the restoration of seagrass meadows and their associated biodiversity thus contributing to their conservation and to people's well-being. A mixed methods approach was used, and results presented in five chapters. Chapter I is a general introduction to the study outlining relevant information, research objectives and hypotheses. Chapter II investigates coastal community’s perceptions on the role of seagrass ecosystems, local impacts of coastal erosion and the need for seagrass restoration in Mozambique. It showed that there is a high awareness of the studied communities to seagrass importance for their livelihoods. Communities supported restoration and identified community engagement as a fundamental component for a successful implementation of restoration projects in their area. Chapter III conducted a mapping inventory for seagrass using open access data sources and investigated change detection in seagrass extent in southern Mozambique using Satellite imagery. In addition, Unmanned Aerial Vehicles (drones) were used to map seagrass to the species level at Inhaca Island. The results showed that seagrass habitats have been losing extent in recent years mainly in Bairro dos Pescadores at northwest Maputo Bay, due to sedimentation from flooding and inappropriate gleaning / clam collection practices. Around Inhaca Island seagrass extent has changed in response to natural causes and from anthropogenic disturbance near the main village and port area. For the first time for this area drones were used to rapidly and effectively acquire high- resolution data for detailed seagrass mapping and monitoring. The different mapping approaches were critically analyzed, and a way forward discussed. In Chapter IV, two different field experiments were conducted simultaneously. One, mimicked disturbance of meadows in terms of digging to collect clams, a common practice in northeastern Maputo Bay to assess the time of natural recovery and another replanted Zostera capensis using different plug sizes. The results indicated that once anthropogenic impacts are ceased, Z. capensis meadows are restored 12 months after impact exclusion. There was a 75% survival rate for the mechanical plug restoration approach; positive implications for seagrass transplantation projects that are both time consuming and financially costly. It was clear from the study that a combination of both impact exclusion and active restoration is needed for seagrass survival. Chapter V investigated the influence of seagrass restoration on the abundance and diversity of macroinvertebrates. After two years of observations, replanted seagrass areas had not recovered compared to natural established seagrass beds. However, a strong linear relationship existed between the abundance of seagrasses and the quantity of associated macrofauna species. Chapter VI provided recommendations on better ways to perform seagrass restoration and management. The main findings of this thesis provide new knowledge on the mapping of seagrass on a local to a regional scale as well as on better restoration practice of the seagrass Zostera capensis, an IUCN Red List vulnerable species. Communities’ perceptions on seagrass were investigated contributing to an understanding of the socio-ecological conservation and management of these critical and yet threatened ecosystems.Thesis (PhD) -- Faculty of Science, School of Environmental Sciences, 202
A comparison of implementation platforms for the visualisation of animal family trees
Get PDFGenealogy is the study of family history. Family trees are used to show ancestry and visualise family history. Animal family trees are different from human family trees as animals have more offspring to represent in a family tree visualisation. Auctioneering organisations, such as Boere Korporasie Beperk (BKB), provide livestock auction catalogues containing pictures of the animal on sale, the animal’s family tree and its breeding and selection data. Modern-day farming has become data-driven and livestock farmers use various online devices and platforms to obtain information, such as real-time milk production, animal health monitoring and to manage farming operations. This study investigated and compared two Business Intelligence (BI) platforms namely Microsoft Power BI and Tableau (Salesforce) and the Python programming language used in the implementation of cattle family tree charts. Animal family tree visualisation requirements were identified from analysing data collected from 23 agriculture users and auction attendees who responded to an online questionnaire. The results of an online survey showed that agriculture users preferred an animal family tree that resembled a human one, which is not currently used in livestock auction catalogues. A conference paper was published based on the survey results. The Design Science Research Methodology (DSRM) was used to aid in creating animal family tree charts using Power BI, Tableau and Python. The author compared the visualisation tools against selected criteria, such as learnability, portability interoperability and security. Usability evaluations using eye tracking were conducted with agriculture users in a usability lab to compare the artefacts developed using Power BI and Python. Tableau was discarded during the implementation process as it did not produce the required family tree visualisation The Technology Acceptance Model (TAM) theory, which seeks to predict the acceptance and use of technology based on users' perception of its usefulness and ease of use, was used to guide the research study in evaluating the artefacts. According to TAM, the adoption of the proposed technology to solve the problem of a static animal family tree in livestock auction catalogues was dependent on the agriculture user’s beliefs. This was based upon that the technology would help them make better buying decisions at livestock auctions effortlessly. The other theory used in this study was the Task Technology Fit (TTF). This theory was used mainly to create the task list to be used in the usability test. The results showed that the author of this work and the agriculture users preferred the artefact produced by Power BI. The learnability and development time was shorter and the User Interface (UI) created was more intuitive. The findings of this study indicated that the present auction catalogue could be supplemented using interactive online animal family tree visualisations created using Power BI. This study recommended that livestock auctioneering organisations should, in addition to providing paper catalogues, provide farmers with an online platform to view the family trees of cattle on auction to enhance purchasing decisions.Thesis (MCom) -- Faculty of Science, School of Computer Science, Mathematics, Physics and Statistics, 202
Exploring the psychological effects of children exposed to domestic violence: a scoping review
Get PDFAn intergenerational sequence of violence exists within families such that children’s direct or indirect experience to domestic violence may result in psychological and behavioural effects in their adulthood. In addition to the implications on children’s emotional and psychological well-being, exposure to domestic violence seems detrimental to children’s expression of emotions. This study aimed to synthesize the type of evidence available on the psychological and emotional effects of domestic violence on children exposed to such experiences. By means of a scoping methodological framework, previously published literature addressing the psychological trauma that children experience due to the exposure to domestic violence was identified. Findings indicate that the consequences on children who have been exposed to domestic violence are broad and can be determined by various factors such as the child’s temperament, environment and genetic predispositions. In this study, 218 studies were screened, with 24 studies assessed for eligibility, whereby only 18 studies met the inclusion criteria for the study. in addition, through the discussion of themes below, it is evident that various factors such as age, cultural differences, gender, etc., play a significant role on how various children can be psychologically affected from exposure to domestic violence. While previous studies have focused on the psychological impacts of domestic violence on the direct victims, the study highlights the gap and attention that should be given to these children, and further encourages limited physical, emotional and any other form of violence that can occur and be potentially harmful to the children exposed to the domestic violence.Thesis (MA) -- Faculty of Health Science, School Behavioural & Lifestyle Sciences, 202
Augmenting the Moore-Penrose generalised Inverse to train neural networks
Get PDFAn Extreme Learning Machine (ELM) is a non-iterative and fast feedforward neural network training algorithm which uses the Moore-Penrose generalised inverse of a matrix to compute the weights of the output layer of the neural network, using a random initialisation for the hidden layer. While ELM has been used to train feedforward neural networks, the effectiveness of the MP generalised to train recurrent neural networks is yet to be investigated. The primary aim of this research was to investigate how biases in the output layer and the MP generalised inverse can be used to train recurrent neural networks. To accomplish this, the Bias Augmented ELM (BA-ELM), which concatenated the hidden layer output matrix with a ones-column vector to simulate the biases in the output layer, was proposed. A variety of datasets generated from optimisation test functions, as well as using real-world regression and classification datasets, were used to validate BA-ELM. The results showed in specific circumstances that BA-ELM was able to perform better than ELM. Following this, Recurrent ELM (R-ELM) was proposed which uses a recurrent hidden layer instead of a feedforward hidden layer. Recurrent neural networks also rely on having functional feedback connections in the recurrent layer. A hybrid training algorithm, Recurrent Hybrid ELM (R-HELM), was proposed, which uses a gradient-based algorithm to optimise the recurrent layer and the MP generalised inverse to compute the output weights. The evaluation of R-ELM and R-HELM algorithms were carried out using three different recurrent architectures on two recurrent tasks derived from the Susceptible- Exposed-Infected-Removed (SEIR) epidemiology model. Various training hyperparameters were evaluated through hyperparameter investigations to investigate their effectiveness on the hybrid training algorithm. With optimal hyperparameters, the hybrid training algorithm was able to achieve better performance than the conventional gradient-based algorithm.Thesis (MSc) -- Faculty of Science, School of Computer Science, Mathematics, Physics and Statistics, 202
Exploring the views of selected, unafflicted women from Walvis bay and Windhoek, Namibia, regarding intimate partner killing and its effect on their well-being.
Get PDFIntimate Partner Killing (IPK), popularly known as passion killing in certain regions, such as in southern parts of Africa, Europe, and America, and also referred to as intimate partner homicide, is a global concern. In Namibia, the rate at which men take the lives of their intimate partners due to failed romantic relationships has sparked apprehension within the communities. This study delves into the perspectives of unafflicted women in Namibia, shedding light on the broader impact of intimate partner killing on female well-being. Using a qualitative approach, the study conducted face-to-face, semi-structured interviews with 14 participants. Grounded theory was applied to analyse the collected data and form categories from it. The data were transcribed and analysed using codes, thematic analysis, and interpretation. The theoretical frameworks intersectionality theory, social feminist theory, structuralist theory and systems theory underpin this study. The findings reveal that intimate partner killing severely affects women’s mental health, leading to vicarious trauma characterised by fear and anxiety. Participants reported difficulty experiencing positive emotions in their romantic relationships due to the pervasive uncertainty of this growing issue. Intimate partner killing leaves women feeling scared, unsafe, and traumatised, often prompting them to exercise extreme caution in their relationships to avoid becoming victims themselves. Symptoms such as feeling unsafe, scared, and traumatised align with post-traumatic stress and generalised anxiety. The study recommends implementing policy interventions and stringent laws in Namibia to prevent or discourage men from resorting to violence against their loved ones. Moreover, establishing assistance centres staffed with social work and mental health professionals is crucial to counselling and supporting women experiencing post-traumatic stress and anxiety resulting from intimate partner violence.Thesis (PhD) -- Faculty of Health Sciences, School of behavioural Sciences, 202
An in vitro evaluation of the anti-breast cancer activity of Nigella sativa extracts and its bioactive compound in combination with curcumin
Get PDFBreast cancer constitutes 23% of all cancers in South African females. Curcumin and Nigella sativa have anti-cancer, anti-metastatic and antioxidant-properties and may be effective against breast cancer. This study focused on the effect of N. sativa extracts or thymoquinone and curcumin, individually and in combination, on breast cancer cells. An MTT assay showed that curcumin reduced cell viability by 50% (IC50) at 18 ± 2.63 μg/mL and thymoquinone (TQ) at 5 ± 0.95 μg/mL against the MDA-MB-231 cells. The IC50 values for curcumin and TQ were 35 ± 6.98 μg/mL and 4 ± 0.96 μg/mL against the MCF-7 cells, respectively. The IC50 value for the NSBE was determined to be 350 ± 55 μg/mL. The IC50 value of NSAE did not fall within the selected concentration range. Synergism was noted for combinations of NSBE with curcumin, and combinations of TQ with curcumin, against both MCF-7 and MDA-MB-231 cells. Two synergistic combinations per treatment per cell line, as determined by the combination index analysis, were chosen for further investigation. The combinations and individual treatments tested against the MCF-10A cells, were not significant, except for NSBE80:CURC20 combination. Curcumin had the most significant anti-oxidant activity; however, no link was noted between the anti-oxidant activity and the cytotoxicity of the combinations. The combination treatments induced apoptosis more effectively than the individual treatments. Caspase-3 dependent apoptosis was noted for NSBE10:CURC90 and TQ80:CURC20 combinations against the MDA-MB-231 cells, and the TQ60:CURC40 combination against the MCF-7 cells. The individual and combined treatments effectively reduced MDA-MB-231 cell adhesion to fibronectin, but not all reduced the cell adhesion to laminin. Based on these results, the combinations of curcumin with TQ or NSBE, have promising anticancer benefits against breast cancer.Thesis (MSc) -- Faculty of Science, School of Biomolecular & Chemical Sciences, 202
An in vitro assessment of pharmacodynamic interactions between selected herbal extracts and anticancer chemotherapeutic agents
Get PDFCancer is a disease caused by uncontrolled growth and spreading of abnormal cells to distant body parts. Approximately 19.3 million and 10 million new cancer cases and cancer deaths, respectively were recorded in 2020. Despite advancements in prevention and therapeutic strategies, cancer remains a global health problem. Cancer patients are increasingly seeking complementary and alternative medicines, of which herbal medicines are the most common form of CAM used by patients. Numerous in vitro and in vivo studies of herbal medicines in cancer have shown that they have antioxidant and anticancer (antitumor and anti-proliferative) properties. While concurrent use with conventional cancer treatments may improve treatment efficacy, alleviate chemotherapy-related side effects, boost the immune system, or impede drug resistance, it can also prompt drug-herbal interactions, and this may affect the pharmacodynamics and pharmacokinetics of the chemotherapeutic drug. This study aimed to determine relevant pharmacodynamic interactions with chemotherapeutic drugs and investigate such interactions' mechanisms. The objectives of the study were to screen the cytotoxicity effects of drug compounds, herbal extracts, and drug-herb combinations, to determine the ability of treatments to induce apoptosis, and to determine the most beneficial treatment. The cytotoxic effects of cisplatin in HeLa cells, tamoxifen in MCF7 cells, 5-fluorouracil in Caco-2 cells, and grapeseed, green tea, fermented rooibos, and green rooibos in all the listed cell lines were evaluated individually and in combination using the bis-Benzamide H 33342 trihydrochloride/propidium iodide (Hoechst 33342/PI) dual staining method. CompuSyn 1.0 Software was used to quantify synergism and antagonism. The mechanism of apoptosis induction of the different synergetic combinations, drug compounds, and herbal extracts was illustrated by quantitative fluorescence image analysis, specifically cell cycle analysis, phosphatidylserine translocation, mitochondrial membrane potential analysis, caspase 3 activation, and reactive oxygen species production using the relevant contrast dyes. Grapeseed displayed cytotoxicity towards MCF7 and HeLa cells (IC50 57.98 and 83.28 μg/mL, respectively). Green tea was only cytotoxic against HeLa cells at an IC50 of 91,92 μg/mL. None of the extracts displayed cytotoxicity against Caco-2 cells (IC50 values > 200μg/mL). If the herbal extracts had inconclusive IC50 values in the three cell lines subsequent experiments were conducted using 100 μg/mL. The results showed that most of the combinations were antagonistic but, some combinations had synergistic or enhancement effects(1:3 for cisplatin with grape seed or green tea and tamoxifen with grapeseed or green tea, and 3:1 for cisplatin with green tea as well as for tamoxifen with fermented rooibos or green rooibos, and all the 1:1 combinations of 5- fluorouracil with all herbal extracts) with a combination index (CI) < 1. Grape seed and green tea were found to induce apoptosis in the three cell lines. Assays that were conducted to detect apoptosis induction showed positive staining for phosphatidylserine (PS), activated caspase 3, and reactive oxygen species (ROS), mitochondrial membrane depolarization. Analysis of the cell cycle showed two things; 1) that grape seed and green tea were apoptotic in HeLa and MCF7 cells only, and 2) the extracts of green tea and green rooibos, and the combinations of grape seed with all drug compounds arrested the cell in more than one phase of the cell cycle. The combinations of grapeseed and green tea potentially induced apoptosis in various manners but those with both rooibos extracts were unclear in all the cell lines. The results for combinations with grapeseed and green tea are promising and provide a basis for further research as combinations of chemotherapeutic drugs and herbal extracts may be effective therapeutic strategies.Thesis (MPharm) -- Faculty of Health Sciences, School of Clinical Care & Medicinal Sciences, 202
The effectiveness of implementing mediation resolutions in Volkswagen South Africa (VWSA) in Kariega 1994-2021
Get PDFThis study aimed to generate an in-depth understanding of how the resolutions are being implemented post-mediation process and the labour union, which in this study is the National Union of Metalworkers of South Africa (NUMSA) and the mediator in the process of managing collective disputes between the employer and employee utilising mediation. The study was limited to the Volkswagen South Africa (VWSA) Kariega plant, therefore, the findings are contextually limited to the VWSA Kariega plant. The theoretical frameworks for this study were relative deprivation theory, frustration-aggression theory, Marxism, and dysfunctional and functional social conflict. These were utilised to explore and describe the phenomenon under investigation. This study employed a qualitative method to answer the research questions and address its problem statement. The data were collected through focus group discussions and elite interviews. The study used a purposive sampling technique, and the sample included 33 participants. The primary research question was: Were the agreements mediated at the VWSA Kariega plant effectively implemented between 1994 and 2021? The study’s findings imply that for the effective implementation of mediation resolutions, the following should be considered: implementing resolutions within the given timeframe, parties at mediation adapting and accommodating new technological changes and disasters at the National Bargaining Forum (NBF) level, prioritising the education and training of shop stewards and VWSA management representatives on NBF agreements, including the implementation thereof. Managing pertinent issues concerning NUMSA and the Automobile Manufacturers Employers Organisation (AMEO), interpreting NBF agreements, and ensuring that parties differentiate between the Labour Relations Act No. 66 of 1995 (LRA) and NBF processes. At the policy level, the findings imply that parties need to understand the LRA’s role and the process of resolving conflict utilising the formal processes stipulated in the LRA and why the LRA cannot override the NBF agreement as an informal process of managing conflict.Thesis (MA) -- Faculty of Humanities, School of Governmental and Social Sciences, 202