This research project investigated the option of steelmaking with the ESS furnace by using computational modelling to estimate steady state decarburisation rates. It focused on understanding metallurgical phenomena that would dictate refining rates of molten pig iron with iron ore. The results obtained are aimed at designers and potential users of the furnace technology to improve their understanding of the expected steady-state process behaviour. A mass-and-energy-balance model with a decarburisation sub-model was developed to estimate feed material requirements for steady state operation. Modelling and simulation results suggest that it may not be possible to produce steel under the conditions proposed. However, the furnace still holds potential if ideal operational conditions are understood and applied. Modelling also gave insight into which areas areas of concern, such as bubble formation in the furnace’s channel induction heaters, and necessity for a well designed refractory lining to contain heat and allow the process to operate at a stable condition under the conditions proposed. Keywords: ESS furnace, steelmaking, metallurgical analysis, modelling, mass and energy balance, decarburisation kinetics / Dissertation (MEng)--University of Pretoria, 2020. / Materials Science and Metallurgical Engineering / MEng / Unrestricted

No description available.

<p> Using the technique of perturbed angular correlations, the rotation of the 4 + 366 keV 152Sm level and the rotation of the 2 + 344 keV 152Gd level was observed. These nuclei were the daughter nuclei of 152Eu, present in the europium - gadolinium and europium - holmium alloys which were commercially obtained (europium concentration was about 1% in both alloys). Anomolously low fields were obtained for 152Gd in gadolinium and in the holmium which raised doubts as to the homogeneity of the alloys.</p> <p> Samples of the alloys were analysed metallurgically. After chemical treatment and microscopic examination the presence of inclusions was detected. Electron microprobe examinations indicated that the inclusions were europium. The metallurgical analysis confirmed the doubts raised by the nuclear experiments showing that the two methods can be employed complementarily for alloy analysis.</p> / Thesis / Master of Science (MSc)

Thesis (MScIng)--University of Stellenbosch, 2007. / ENGLISH ABSTRACT: Due to economic and environmental constraints mining operations are placed under increasing pressure to effectively manage and operate tailings disposal operations. Restrictions imposed on water usage and tailings operations footprint have led to higher density and wider particle size distribution slurries conveyed to tailings areas. One means of efficiently disposing the tailings is co-disposal. In this method a concentrated fine vehicle slurry is used to convey a coarser fraction. This produces a higher density of tailings, with a number of advantages both upstream and downstream of the tailings process. Limited research has been conducted on the effect of coarse particles on the non-Newtonian rheological properties of these slurries. This lack of information complicates the design and reliable operation of these systems. This project aims at gaining a clearer understanding as to the mechanisms involved in the addition of coarse particles to a fine clay slurry vehicle; and to provide a means of estimating the measured slurry rheological properties. A number of experiments were designed to test the slurry (both Kaolin only, and Kaolin-coarse particle mixtures) rheological properties using a Couette viscometer (for the dynamic flow properties of yield stress and plastic viscosity) and a vane instrument (for the static yield stress measurements). The slurries were prepared in varying Kaolin clay solids concentrations with reverse osmosis water. Glass beads and two types of industrial sand were used as the coarse fractions. All of the coarse particles had a similar size but varied significantly in shape. Slurry pH and temperature readings were monitored throughout the tests. Tests were done initially on clay only slurries. The rheological properties of these slurries were repeatable, and no noticeable variations of properties with time were observed. The yield stress (both static and dynamic) and plastic viscosity data were well correlated with established relationships. Coarse particles were added to the clay only slurries, and then removed. The remaining clay only slurry exhibited the same rheological properties as the initial clay only slurry. The presence of coarse particles increased all the measured rheological properties (i.e. dynamic yield stress, Bingham viscosity, and static vane yield stress) in a fashion resembling the effect of adding clay to a clay only slurry. In addition, the change in measured rheological property by addition of coarse particle was independent of the clay fraction in the clay slurry. Furthermore, with both the clay only slurries and clay and coarse sand slurries, a constant linear relationship existed between the static and dynamic yield stress. Several correlations from the literature were found to provide reasonable prediction of the rheological property variations observed. These empirical and semi-empirical models however did little to explain the mechanisms involved in coarse particle addition. A new correlation has been proposed, Residual Clay Concentration, which predicts the change in rheological property based on an additional clay concentration, which in turn is a linear function of the coarse particle concentration. The accuracy of this model further strengthens the belief that the coarse particle acts in a similar fashion to a floc. By means of a case study example the importance of selecting an appropriate model for design was illustrated. The Residual Clay Concentration method provided the most conservative results. This combined with its theoretical basis strengthens the models recommendation for use in design. / AFRIKAANSE OPSOMMING: As gevolg van ekonomiese en omgewings beperkinge word mynwese nywerhede onder toenemende druk geplaas om doeltreffende afvalstroom bestuur en operasie toe te pas. Beperkinge geplaas op water gebruik en afvalstroom area-groote ly tot hoër digthede en wyer partikel-grooteverspreidings van flodders vervoer na afval areas. Een manier om van die afval doeltreffend ontslae te raak en te berg is deur medeberging. In die metode word ‘n gekonsentreerde fyn flodder gebruik as draer van ‘n growwer partikel-fraksie. Dit ly tot ‘n hoër digtheid flodder, met verskeie voordele in beide die op – en afstroom prosesse. Beperkte navorsing is gedoen op die effek van growwe partikels op die nie-Newoniese rheolgiese eienskappe van hierdie flodders. Hierdie tekort aan informasie maak die effektiewe, betroubare bedryf en operasie van die sisteme meer ingewikkeld. Hierdie projek is daarheen gemik om ‘n beter begrip te ontwikkel met betrekking tot die meganismes betrokke in die byvoeging van growwe partikels aan ‘n fyn klei-agtige flodder draer; en om ‘n manier te voorsien wat die rheologiese eienskappe kan beraam. Verskeie eksperimente was ontwerp om die flodders (beide slegs Kaolien, en Kaoliengrowwe partikel mengsels) se rheologiese eienskappe te toets deur die gebruik van ‘n Couette-viskometer. Die Couette viskometer was gebruik om die dinamiese eienskappe (van grens-spanning, en plastiese viskositet) te meet. ‘n Vaan apparaat is gebruik om die eienskap van statiese grens-spanning te meet. Die flodders was voorberei in verskeie Kaolien konsentrasies met tru-osmosis water. Glas krale en twee tipes industriële sand is gebruik as die growwe fraksies. Al die growwe partikels het soortgelyke groottes gehad, maar het grootliks verskil in vorm. Die flodder pH en temperatuur lesings is deurentyd nagegaan. Toetse was aanvanklik gedoen op die klei-alleenlike flodders. Die gemete reologiese eienskappe van die flodders was herhaalbaar, en geen opmerkbare veranderinge van die eienskappe met betrekking tot tyd is gemeet nie. Die grens-spanning (beide statiese en dinamiese) en plastiese viskositeit is goed gekorrelleer met gevestigde verhoudinge. Growwe partikels is aan die klei-alleenlike flodders bygevoeg, en daarnae verwyder. Die oorblywende klei-alleenlike flodder het dieselfde gemete rheologiese eienskappe getoon as die oorspronklike klei-allenlike flodder. Die teenwoordigheid van growwe partikels het na ‘n toename van al die gamete rheologiese eienskappe gelei wat fisies baie soortgelyk is aan die byvoeging van klei tot ‘n klei-alleenlike flodder. Verder, met beide die klei-alleenlike en klei-growwe partikel flodders het ‘n konstante liniëre funksie tussen die statiese en dinamiese grens-spannings bestaan. Verskeie verhoudings uit die literatuur het goeie korrelasie bewerkstellig met die waargenome rheologie veranderinge. Hierdie empiriese en semi-empiriese modelle doen egter min om die megansimes betrokke in die toevoeging van growwe partikels te verduidelik. ‘n Nuwe korrelasie is voorgestel, naamlik die Residu Klei Konsentrasie. Hierdie model voorspel die verandering in reologiese eienskappe gebaseer op ‘n addisionele klei konsentrasie, wat ‘n liniëre funkise is van die growwe partikel konsentrasie. Die goeie korrelasie gesien met die model versterk die idée dat die growwe partikel in ‘n soortgelyke manier as ‘n flok gedra in die teenwoordigheid van ander flokke. Deur middel van ‘n tipiese industriële voorbeeld is die belangrikheid in die keuse van die regte korrelasie geillustreer. Die Residu Klei Konsentrasie metode het die mees konservatiewe resultate gelewer. Hierdie feit gekombineerd met die model se soliede teoreitiese beginsels versterk dit as voorgestelde korrelasie vir ontwerp.

The Bushveld Igneous Complex (BIC) in South Africa holds approximately three quarters of the world’s viable chromite ore deposits. Most chromite is used for the production of ferrochromium (FeCr), which is a relatively crude alloy. Several chromite seams exist in the BIC. The most economically viable seams are the lower group 6 (LG6), the middle group 1 and 2 (MG1 and 2) and the upper group 2 (UG2) seams. The LG and MG seams are exploited specifically for their chromium content, whereas the UG2 seam is mainly exploited for platinum group minerals (PGM). However, the upgraded UG2 chromite in the PGM tailing is increasingly being used as a feedstock for FeCr production. Many different process variations exist for FeCr production. However, the process option applied in most green and brown field FeCr expansion projects during the last two decades in South Africa has been the oxidative sintered process (also referred to as the Outotec or Outokumpu process). Notwithstanding the common application of the oxidative sintered pelletised feed technology in the South African FeCr industry, and the increased utilisation of beneficiated UG2 ore, a direct comparison of the physical properties of oxidative sintered pellets produced from UG2 with the physical properties of that produced with conventional metallurgical grade chromite ore is currently lacking in the peer-reviewed scientific literature. Therefore, the physical properties of oxidative sintered pellets produced from a typical beneficiated South African UG2 ore were compared with the physical properties of that produced with conventional South African metallurgical grade chromite ore in this study. The results indicated that the case study metallurgical grade chromite ore required 13 kWh/t more energy to mill than the case study UG2 ore prior to pelletisation, which can lead to substantial cost savings. The compressive strength of the oxidative sintered pellets of both case study ores statistically showed that oxidative sintered pellets made from UG2 ore Abstract were the same or better than those prepared from metallurgical grade chromite ore. A comparison of the abrasion strengths of the oxidative sintered pellets of both case study ores indicated that the pellets prepared from UG2 ore were superior compared to the metallurgical grade pellets. The reasons for the superior UG2 pellet strength were investigated with backscatter-, secondary electron- and elemental X-ray mapping scanning electron microscopy (SEM), which indicated that differences in crystalline structures at least partially contributed to the differences observed. Results presented here can be utilised by FeCr producers to better quantify the advantages and disadvantages associated with the use of UG2 ore for FeCr production. / MSc (Chemistry), North-West University, Potchefstroom Campus, 2014

Ferrochrome (FeCr) is a crude alloy containing chromium (Cr) and iron (Fe). FeCr is mainly used for the production of stainless steel, which is an important modern-day alloy. FeCr is produced from chromite ore through various smelting methods. In this study, the focus was on the pelletised chromite pre-reduction process, which is also referred to as the solid state reduction of chromite. In this process, fine chromite ore, a clay binder and a carbon reductant are dry milled, agglomerated (pelletised) and pre-reduced (solid state reduction) in a rotary kiln. The pre-reduced pellets are then charged hot, immediately after exiting the rotary kiln, into a closed submerged arc furnace (SAF). This production process option has the lowest specific energy consumption (SEC), i.e. MWh/ton FeCr produced, of all the FeCr production processes that are commercially applied. Other advantages associated with the application of the pelletised chromite pre-reduction process are that it eliminates the use of chromite fines, has a high Cr recovery, and produces low sulphur- (S) and silicon (Si)-containing FeCr. The main disadvantage of the pelletised chromite pre-reduction process is that it requires extensive metallurgical control due to the variances in the levels of pre-reduction achieved and carbon content of the pre-reduced pelletised furnace feed material. This implies that the metallurgical carbon balance has to be changed regularly to prevent the process from becoming carbon deficient (also referred to as ‘under coke’) or over carbon (also referred to as ‘over coke’). The analytical technique currently applied to determine the level of chromite pre-reduction is time consuming, making it difficult and expensive to deal with large numbers of samples. In an attempt to develop a technique that would be faster to determine the level of chromite pre-reduction, a new analytical method using a combination of scanning electron microscopy (SEM), image processing and computational techniques was investigated in this study. Metallurgical grade chromite (<1 mm), anthracite breeze (<1 mm), and fine FeCr (<1 mm) that were used to prepare pellets in the laboratory, as well as industrially produced pre-reduced pellets that had already been milled in preparation for the determination of the pre-reduction level with wet chemical analysis were received from a large South African FeCr producer. These laboratory prepared pellets and the industrially produced pellet mixtures were considered in this investigation. Samples were moulded in resin and polished in order to obtain SEM micrographs of the polished cross sections. Elements with higher molecular weights are indicated by lighter greyscale, while elements with lower molecular weights are indicated by darker greyscale in SEM micrographs. This basic principle was applied in the development of the new analytical technique to determine the level of chromite pre-reduction, with the hypothesis that the pixel count of white pixels (representing metallised particles), divided by the combined pixel count of white (representing metallised particles) and grey (representing chromite particles) pixels would be directly related to the level of chromite pre-reduction determined with the current wet chemical method. This hypothesis can be mathematically expressed as: The newly-developed analytical method was validated by correlating the white pixel% calculated with the chromite pre-reduction levels (%) determined with wet chemical analysis of laboratory prepared and industrially produced pellet mixtures, which had R2 values of 0.998 and 0.919, respectively. This suggests that the method can be used to determine chromite pre-reduction accurately. / MSc (Engineering Sciences in Chemical Engineering), North-West University, Potchefstroom Campus, 2015

The Bushveld Igneous Complex (BIC) in South Africa holds approximately three quarters of the world’s viable chromite ore deposits. Most chromite is used for the production of ferrochromium (FeCr), which is a relatively crude alloy. Several chromite seams exist in the BIC. The most economically viable seams are the lower group 6 (LG6), the middle group 1 and 2 (MG1 and 2) and the upper group 2 (UG2) seams. The LG and MG seams are exploited specifically for their chromium content, whereas the UG2 seam is mainly exploited for platinum group minerals (PGM). However, the upgraded UG2 chromite in the PGM tailing is increasingly being used as a feedstock for FeCr production. Many different process variations exist for FeCr production. However, the process option applied in most green and brown field FeCr expansion projects during the last two decades in South Africa has been the oxidative sintered process (also referred to as the Outotec or Outokumpu process). Notwithstanding the common application of the oxidative sintered pelletised feed technology in the South African FeCr industry, and the increased utilisation of beneficiated UG2 ore, a direct comparison of the physical properties of oxidative sintered pellets produced from UG2 with the physical properties of that produced with conventional metallurgical grade chromite ore is currently lacking in the peer-reviewed scientific literature. Therefore, the physical properties of oxidative sintered pellets produced from a typical beneficiated South African UG2 ore were compared with the physical properties of that produced with conventional South African metallurgical grade chromite ore in this study. The results indicated that the case study metallurgical grade chromite ore required 13 kWh/t more energy to mill than the case study UG2 ore prior to pelletisation, which can lead to substantial cost savings. The compressive strength of the oxidative sintered pellets of both case study ores statistically showed that oxidative sintered pellets made from UG2 ore Abstract were the same or better than those prepared from metallurgical grade chromite ore. A comparison of the abrasion strengths of the oxidative sintered pellets of both case study ores indicated that the pellets prepared from UG2 ore were superior compared to the metallurgical grade pellets. The reasons for the superior UG2 pellet strength were investigated with backscatter-, secondary electron- and elemental X-ray mapping scanning electron microscopy (SEM), which indicated that differences in crystalline structures at least partially contributed to the differences observed. Results presented here can be utilised by FeCr producers to better quantify the advantages and disadvantages associated with the use of UG2 ore for FeCr production. / MSc (Chemistry), North-West University, Potchefstroom Campus, 2014

Ferrochrome (FeCr) is a crude alloy containing chromium (Cr) and iron (Fe). FeCr is mainly used for the production of stainless steel, which is an important modern-day alloy. FeCr is produced from chromite ore through various smelting methods. In this study, the focus was on the pelletised chromite pre-reduction process, which is also referred to as the solid state reduction of chromite. In this process, fine chromite ore, a clay binder and a carbon reductant are dry milled, agglomerated (pelletised) and pre-reduced (solid state reduction) in a rotary kiln. The pre-reduced pellets are then charged hot, immediately after exiting the rotary kiln, into a closed submerged arc furnace (SAF). This production process option has the lowest specific energy consumption (SEC), i.e. MWh/ton FeCr produced, of all the FeCr production processes that are commercially applied. Other advantages associated with the application of the pelletised chromite pre-reduction process are that it eliminates the use of chromite fines, has a high Cr recovery, and produces low sulphur- (S) and silicon (Si)-containing FeCr. The main disadvantage of the pelletised chromite pre-reduction process is that it requires extensive metallurgical control due to the variances in the levels of pre-reduction achieved and carbon content of the pre-reduced pelletised furnace feed material. This implies that the metallurgical carbon balance has to be changed regularly to prevent the process from becoming carbon deficient (also referred to as ‘under coke’) or over carbon (also referred to as ‘over coke’). The analytical technique currently applied to determine the level of chromite pre-reduction is time consuming, making it difficult and expensive to deal with large numbers of samples. In an attempt to develop a technique that would be faster to determine the level of chromite pre-reduction, a new analytical method using a combination of scanning electron microscopy (SEM), image processing and computational techniques was investigated in this study. Metallurgical grade chromite (<1 mm), anthracite breeze (<1 mm), and fine FeCr (<1 mm) that were used to prepare pellets in the laboratory, as well as industrially produced pre-reduced pellets that had already been milled in preparation for the determination of the pre-reduction level with wet chemical analysis were received from a large South African FeCr producer. These laboratory prepared pellets and the industrially produced pellet mixtures were considered in this investigation. Samples were moulded in resin and polished in order to obtain SEM micrographs of the polished cross sections. Elements with higher molecular weights are indicated by lighter greyscale, while elements with lower molecular weights are indicated by darker greyscale in SEM micrographs. This basic principle was applied in the development of the new analytical technique to determine the level of chromite pre-reduction, with the hypothesis that the pixel count of white pixels (representing metallised particles), divided by the combined pixel count of white (representing metallised particles) and grey (representing chromite particles) pixels would be directly related to the level of chromite pre-reduction determined with the current wet chemical method. This hypothesis can be mathematically expressed as: The newly-developed analytical method was validated by correlating the white pixel% calculated with the chromite pre-reduction levels (%) determined with wet chemical analysis of laboratory prepared and industrially produced pellet mixtures, which had R2 values of 0.998 and 0.919, respectively. This suggests that the method can be used to determine chromite pre-reduction accurately. / MSc (Engineering Sciences in Chemical Engineering), North-West University, Potchefstroom Campus, 2015

Thesis (PhD (Process Engineering))--University of Stellenbosch, 2007. / Dissertation presented for the degree of Doctor of Philosophy in Engineering at the University of Stellenbosch. / ENGLISH ABSTRACT: The development of advanced methods of process monitoring, diagnosis, and control has been identified as a major 21st century challenge in control systems research and application. This is particularly the case for chemical and metallurgical operations owing to the lack of expressive fundamental models as well as the nonlinear nature of most process systems, which makes established linearization methods unsuitable. As a result, efforts have been directed in the search of alternative approaches that do not require fundamental or analytical models. Data-based methods provide a very promising alternative in this regard, given the huge volumes of data being collected in modern process operations as well as advances in both theoretical and practical aspects of extracting information from observations. In this thesis, the use of kernel-based learning methods in fault detection and diagnosis of complex processes is considered. Kernel-based machine learning methods are a robust family of algorithms founded on insights from statistical learning theory. Instead of estimating a decision function on the basis of minimizing the training error as other learning algorithms, kernel methods use a criterion called large margin maximization to estimate a linear learning rule on data embedded in a suitable feature space. The embedding is implicitly defined by the choice of a kernel function and corresponds to inducing a nonlinear learning rule in the original measurement space. Large margin maximization corresponds to developing an algorithm with theoretical guarantees on how well it will perform on unseen data. In the first contribution, the characterization of time series data from process plants is investigated. Whereas complex processes are difficult to model from first principles, they can be identified using historic process time series data and a suitable model structure. However, prior to fitting such a model, it is important to establish whether the time series data justify the selected model structure. Singular spectrum analysis (SSA) has been used for time series identification. A nonlinear extension of SSA is proposed for classification of time series. Using benchmark systems, the proposed extension is shown to perform better than linear SSA. Moreover, the method is shown to be useful for filtering noise in time series data and, therefore, has potential applications in other tasks such as data rectification and gross error detection. Multivariate statistical process monitoring methods are well-established techniques for efficient information extraction from multivariate data. Such information is usually compact and amenable to graphical representation in two or three dimensional plots. For process monitoring purposes control limits are also plotted on these charts. These control limits are usually based on a hypothesized analytical distribution, typically the Gaussian normal distribution. A robust approach for estimating con dence bounds using the reference data is proposed. The method is based on one-class classification methods. The usefulness of using data to define a confidence bound in reducing fault detection errors is illustrated using plant data. The use of both linear and nonlinear supervised feature extraction is also investigated. The advantages of supervised feature extraction using kernel methods are highlighted via illustrative case studies. A general strategy for fault detection and diagnosis is proposed that integrates feature extraction methods, fault identification, and different methods to estimate confidence bounds. For kernel-based approaches, the general framework allows for interpretation of the results in the input space instead of the feature space. An important step in process monitoring is identifying a variable responsible for a fault. Although all faults that can occur at any plant cannot be known beforehand, it is possible to use knowledge of previous faults or simulations to anticipate their recurrence. A framework for fault diagnosis using one-class support vector machine (SVM) classification is proposed. Compared to other previously studied techniques, the one-class SVM approach is shown to have generally better robustness and performance characteristics. Most methods for process monitoring make little use of data collected under normal operating conditions, whereas most quality issues in process plants are known to occur when the process is in-control . In the final contribution, a methodology for continuous optimization of process performance is proposed that combines support vector learning with decision trees. The methodology is based on continuous search for quality improvements by challenging the normal operating condition regions established via statistical control. Simulated and plant data are used to illustrate the approach. / AFRIKAANSE OPSOMMING: Die ontwikkeling van gevorderde metodes van prosesmonitering, diagnose en -beheer is geïdentifiseer as 'n groot 21ste eeuse uitdaging in die navorsing en toepassing van beheerstelsels. Dit is veral die geval in die chemiese en metallurgiese bedryf, a.g.v. die gebrek aan fundamentele modelle, sowel as die nielineêre aard van meeste prosesstelsels, wat gevestigde benaderings tot linearisasie ongeskik maak. Die gevolg is dat pogings aangewend word om te soek na alternatiewe benaderings wat nie fundamentele of analitiese modelle benodig nie. Data-gebaseerde metodes voorsien belowende alternatiewe in dié verband, gegewe die enorme volumes data wat in moderne prosesaanlegte geberg word, sowel as die vooruitgang wat gemaak word in beide die teoretiese en praktiese aspekte van die onttrekking van inligting uit waarnemings. In die tesis word die gebruik van kern-gebaseerde metodes vir foutopsporing en -diagnose van komplekse prosesse beskou. Kern-gebaseerde masjienleermetodes is 'n robuuste familie van metodes gefundeer op insigte uit statistiese leerteorie. Instede daarvan om 'n besluitnemingsfunksie te beraam deur passingsfoute op verwysingsdata te minimeer, soos wat gedoen word met ander leermetodes, gebruik kern-metodes 'n kriterium genaamd groot marge maksimering om lineêre reëls te pas op data wat ingebed is in 'n geskikte kenmerkruimte. Die inbedding word implisiet gedefinieer deur die keuse van die kern-funksie en stem ooreen met die indusering van 'n nielineêre reël in die oorspronklike meetruimte. Groot marge-maksimering stem ooreen met die ontwikkeling van algoritmes waarvan die prestasie t.o.v. die passing van nuwe data teoreties gewaarborg is. In die eerste bydrae word die karakterisering van tydreeksdata van prosesaanlegte ondersoek. Alhoewel komplekse prosesse moeilik is om vanaf eerste beginsels te modelleer, kan hulle geïdentifiseer word uit historiese tydreeksdata en geskikte modelstrukture. Voor so 'n model gepas word, is dit belangrik om vas te stel of die tydreeksdata wel die geselekteerde modelstruktuur ondersteun. 'n Nielineêre uitbreiding van singuliere spektrale analise (SSA) is voorgestel vir die klassifikasie van tydreekse. Deur gebruik te maak van geykte stelsels, is aangetoon dat die voorgestelde uitbreiding beter presteer as lineêre SSA. Tewens, daar word ook aangetoon dat die metode nuttig is vir die verwydering van geraas in tydreeksdata en daarom ook potensiële toepassings het in ander take, soos datarektifikasie en die opsporing van sistematiese foute in data. Meerveranderlike statistiese prosesmonitering is goed gevestig vir die doeltreffende onttrekking van inligting uit meerveranderlike data. Sulke inligting is gewoonlik kompak en geskik vir voorstelling in twee- of drie-dimensionele grafieke. Vir die doeleindes van prosesmonitering word beheerlimiete dikwels op sulke grafieke aangestip. Hierdie beheerlimiete word gewoonlik gebaseer op 'n hipotetiese analitiese verspreiding van die data, tipiese gebaseer op 'n Gaussiaanse model. 'n Robuuste benadering vir die beraming van betroubaarheidslimiete gebaseer op verwysingsdata, word in die tesis voorgestel. Die metode is gebaseer op eenklas-klassifikasie en die nut daarvan deur data te gebruik om die betroubaarheidsgrense te beraam ten einde foutopsporing te optimeer, word geïllustreer aan die hand van aanlegdata. Die gebruik van beide lineêre en nielineêre oorsiggedrewe kenmerkonttrekking is vervolgens ondersoek. Die voordele van oorsiggedrewe kenmerkonttrekking deur van kern-metodes gebruik te maak is beklemtoon deur middel van illustratiewe gevallestudies. 'n Algemene strategie vir foutopsporing en -diagnose word voorgestel, wat kenmerkonttrekkingsmetodes, foutidenti kasie en verskillende metodes om betroubaarheidsgrense te beraam saamsnoer. Vir kern-gebaseerde metodes laat die algemene raamwerk toe dat die resultate in die invoerruimte vertolk kan word, in plaas van in die kenmerkruimte. 'n Belangrike stap in prosesmonitering is om veranderlikes te identifiseer wat verantwoordelik is vir foute. Alhoewel alle foute wat by 'n chemiese aanleg kan plaasvind, nie vooraf bekend kan wees nie, is dit moontlik om kennis van vorige foute of simulasies te gebruik om die herhaalde voorkoms van die foute te antisipeer. 'n Raamwerk vir foutdiagnose wat van eenklas-steunvektormasjiene (SVM) gebruik maak is voorgestel. Vergeleke met ander tegnieke wat voorheen bestudeer is, is aangetoon dat die eenklas-SVM benadering oor die algemeen beter robuustheid en prestasiekenmerke het. Meeste metodes vir prosesmonitering maak min gebruik van data wat opgeneem is onder normale bedryfstoestande, alhoewel meeste kwaliteitsprobleme ondervind word waneer die proses onder beheer is. In die laaste bydrae, is 'n metodologie vir die kontinue optimering van prosesprestasie voorgestel, wat steunvektormasjiene en beslissingsbome kombineer. Die metodologie is gebaseer op die kontinue soeke na kwaliteitsverbeteringe deur die normale bedryfstoestandsgrense, soos bepaal deur statistiese beheer, te toets. Gesimuleerde en werklike aanlegdata is gebruik om die benadering te illustreer.

Ce travail étudie le potentiel du silicium de type n purifié par voie métallurgique pour la fabrication de cellules photovoltaïques à bas coût. Les teneurs élevées en dopants conduisent à des gammes de résistivités larges et faibles, ainsi qu’à une diminution de la durée de vie des porteurs de charge.La fabrication de cellules photovoltaïques a permis d’obtenir des rendements de conversion variant de 13.7% à 15.0% sur 148.6cm². Avec un procédé de fabrication amélioré, des rendements de 16.0% pourraient être obtenus. La résistivité des plaquettes a été identifiée comme facteur limitant les performances des cellules. Le co-dopage au gallium a été proposé pour augmenter la gamme de résistivité.Les cellules photovoltaïques réalisées montrent une excellente stabilité sous illumination et de faibles coefficients en température de la tension de circuit-ouvert. Ces travaux de thèse ont permis de définir le potentiel du silicium de type n purifié par voie métallurgique et de définir les spécifications nécessaires initiales au niveau de la charge à purifier pour permettre la fabrication de cellules photovoltaïques efficaces. / This work studies the potential of n-type silicon purified via the metallurgical route for the fabrication of low cost photovoltaic cells. The high level of doping species leads to a large range of low resistivity, as well as reduced carriers’ lifetime. The fabrication of photovoltaic cells led to conversion efficiencies varying from 13.7% to 15.0% on 148.6 cm². With an improved fabrication process, efficiencies of 16.0% could be obtained. The resistivity has been identified as the limiting factor on the cells’ efficiency. Gallium co-doping has been proposed in order to increase the resistivity range.The fabricated photovoltaic cells show an excellent stability under illumination with weak temperature coefficients of the open circuit voltage. This PhD work led to the knowledge of the potential of n-type silicon purified via the metallurgical route, and to define the charge specifications required to the fabrication of efficient photovoltaic cells.