Thesis (PhD (Process Engineering))--University of Stellenbosch, 2006. / In the smelting of copper waste slags to recover cobalt and copper, the prediction of the metal liquidus temperature and the associated superheat for liquid metal handling for subsequent treatments cannot be done with certainty, making the management of furnace integrity very difficult. By studying the phase equilibria and solution thermodynamics in liquid ferrocobalt new experimental data on the liquidus temperature and phase equilibria of the quaternary system can contribute to the improvement of existing copper slag smelting processes. This will alleviate the operational uncertainties and difficulties associated with ferrocobalt production in electric arc furnaces. There is no specific literature available that describes the physicochemical and thermochemical properties of the ferrocobalt produced from smelting of waste copper slags. Therefore, the quaternary system Fe-Co-Cu-Si has been characterised by studying and reviewing the binary and ternary subsystems with respect to the high temperature phase equilibria. The ferrocobalt metal has been modelled on the Fe-Co-Cu-Si quaternary system. The liquidus temperatures and phase equilibria in the Fe-Co-Cu-Si system, within the composition and temperature regimes pertinent to smelting of slag, were investigated by differential thermal analysis and metallography. Drop-quench techniques coupled with scanning electron microscopy were used to study the phase equilibria. The activity of silicon in liquid Fe-Co-Cu-Si at 1450 °C was calculated from gas-alloy-silica equilibrium experiments conducted in controlled oxygen partial pressure atmospheres at 10-13 P, 10P -14 P, and 10P -15 atmosphere (absolute) corresponding to the conditions found in the industrial application. The liquidus temperature of the quaternary Fe-Co-Cu-Si is influenced by the content of silicon in the system. When silicon is added to the Fe-Co-Cu ternary the liquidus temperature is lowered in the new system (Fe-Co-Cu-Si). In the range of silicon content studied (0 0.1) < XSi ≤ , the liquidus temperature decreased by over 70 °C. The liquidus temperature profiles of the subsystems of the quaternary Fe-Co-Cu-Si, show large composition dependence too, except in the Fe-Co system. In the ternary Fe-Co-Cu, the liquidus temperature decreases with increasing copper content and is characteristic of the profiles of the liquidus lines in the binary subsystems Fe-Cu and Co-Cu.In the dilute concentrations of silicon, it is shown that the phase equilibria in the quaternary system have attributes of the Fe-Si and Fe-Cu-Si systems. Silicon is associated more with the iron rich phase than it is with the copper rich phase. It stabilises the metastable liquid immiscibility when added to the Fe-Cu, Co-Cu, and Fe-Co-Cu in the corresponding ternary systems Fe-Cu-Si, Co-Cu-Si and quaternary Fe-Co-Cu-Si system. The activity of silicon in liquid Fe-Co-Cu-Si at 1450 °C, in the composition range 1 to 5 wt. %Si exhibits a negative deviation from ideal liquid solution behaviour. The activity coefficient approaches a constant value of 0.2×10P -3 P, with pure liquid silicon as reference state, as the silicon concentration approaches zero implying a Henrian solution behaviour. This information should be useful in the thermodynamic modelling of the system.

Thesis (PhD)--University of Stellenbosch, 2004. / ENGLISH ABSTRACT: The objective of this work was to achieve a completely automated fire assay system for the analysis of process control samples on a flotation plant in less than 120 minutes. With this in mind, a systems engineering approach was undertaken. The physical and chemical characteristics of the technology for each subsystem were investigated in turn and the critical factors that influenced accuracy, precision and analysis time were identified and optimised. Some of the key developments achieved during this work were: · Existing technology for the sampling, filtering, drying and grinding of flotation plant samples were evaluated and where necessary, modified for this application. · The fusion system was totally re-designed with a bottom-loading configuration called FIFA (Fast In-line Fire Assay) to make automation with a central robot possible. With the fast fusing flux developed, a quantitative collection of the platinum group elements with a fifteen-minute fusion was achieved compared to an hour for the classical method. · A robust automated separator system was developed to isolate the lead collector from the fusion in the molten state thereby separating it quantitatively from the slag. This allowed the automation of the entire fire assay process. · Methods to prepare lead standards for calibration were developed. These were used to optimise analytical protocols for the analysis of platinum group elements in lead using a spark optical emission spectrometer. This made it possible to accurately determine the quantities of platinum group elements in lead samples prepared by the automated fire assay system. · A fully automated system was developed that could meet the accuracy and precision requirements for the analysis of tailings and feed grade samples in concentrator slurry streams in less than one hour compared with the 24-72 hours required when using classical methods. The new fire assay technology including flux, FIFA system, oxygen lance and separator were all patented along with the automation vendor. This technology has made the first fully automated fire assay system a reality. / AFRIKAANSE OPSOMMING: Die hoofoogmerk van die studie was om ‘n totale geoutomatiseerde vuuressaieering stelsel te ontwerp vir die ontleding van prosesbeheermonsters van ‘n flotasieaanleg in ‘n bestek van 120 minute. Gedurende die ontwerp is ‘n ingeneursstelselbenadering gebruik. Die fisiese en chemiese kenmerke van elke deel van die tegnologie is eers afsonderlik en dan as ‘n geheel ondersoek. Die bepalende faktore wat akkuraatheid, presisie en ontledingstyd beinvloed het was geidentifiseer en geoptimeer. Die hoofpunte van die werk behels onder andere die volgende: · Bestaande tegnologie vir monsterneming, filtrasie, droging en vermaling van flotasiemonsters was ondersoek en is, waar nodig, aangepas vir die finale stelsel. · Die smeltingsisteem was in geheel herontwerp om monsters van onder in die FIFA (Fast In-line Fire Assay) sisteem te laai en sodoende die outomatisering met ‘n sentrale robot te vergemaklik. ‘n Vinnig smeltende vloeimiddel was ontwikkel wat ‘n kwantitatiewe versameling van die platinum groep elemente binne ‘n tydsduur van vyftien minute moontlik gemaak het, in vergelyke met die oorspronklike duur van die klassieke smelt metode van een uur. · ‘n Outomatiese skeier was ontwikkel waarmee die gesmelte loodversamelaar geskei kon word van die slakfase. Met die nuwe stelsel kon die hele vuuressaieerproses outomaties verloop. · Metodes is ontwikkel om loodstandaarde vir kalibrasie doeleindes te berei. Die standaarde is op hulle beurt weer gebruik om ‘n ontleding protokol daar te stel vir die analiese van die platinum groep elemente in lood, met behulp van ‘n vonkontlading-optiese-uitstraling-spektrometriese instrument. Ten einde was dit moontlik om outomaties klein hoeveelhede van die platinum groep elemente in monsters akkuraat te bepaal, na voorbereiding met behulp van die geoutomatiseerde vuuressaieering stelsel. · Die volle geoutimatiseerde stelsel was ontwikkel wat aan die akkurate en noukeurige vereistes voldoen het vir die ontleding van flotasie-uitskot-envoergraad monsters van die konsentraataanleg binne die bestek van ‘n uur. Die nuwe vuuressaieer tegnologie, insluitend die vinnig smeltende vloeimiddel, FIFA en skeier stelsels, asook die suurstof lanset is gepatenteer met die vervaardiger. Die studie het gelei tot die eerste volle geoutomatiseerde vuuressaieering stelsel wat tans gebruik word in die industrie.

Thesis (MScIng)--University of Stellenbosch, 2004. / ENGLISH ABSTRACT: This thesis details results obtained in an experimental study conducted to determine the effects of operating temperature, oxygen partial pressure, bulk chromium oxide content and bulk FeOx/MgO ratio on the solubility of chromium oxide in melter type slags in the platinum industry. Two PGM-containing layers in the Bushveld Complex in South Africa, the Merensky and UG2 reefs, are currently being mined for the extraction of base metals and platinum group metals (PGM). While the Merensky reef is a pyroxenitic layer, the UG2 reef is a platiniferous chromitite seam. Due to a gradual depletion in Merensky ore reserves, platinum producers have been moving towards the processing of more UG2 concentrates, which are higher in chromium oxide content. The technical difficulties associated with the smelting of concentrates with high chromium oxide contents is a matter of concern. The formation of chromite spinels in melts increases liquidus temperatures and viscosities and subsequently hampers tapping of slags and mattes from furnaces. Bottom build-up from the smelting of high chromium oxide containing concentrates could reduce effective furnace volume. From the literature reviewed it was found that very few published investigations covered melt compositions and oxygen partial pressures similar to those encountered in the platinum industry. Relevant studies were found to deal with significantly lower bulk chromium oxide and iron oxide contents. It became clear that a need exists for information on the behaviour of chromium oxide and its effects on phase chemistry and stability in melter slags. It was decided to study the phase equilibria through drop-quench experiments using six synthetic slags with bulk FeOx/MgO ratios between 0.6 and 1.9 and bulk chromium oxide contents between 1.2 and 7 wt%. Temperatures investigated were 1400, 1500 and 1600°C. The oxygen partial pressure was varied between 6.8x10-10 atm at 1400°C to8.3x10-5 atm at 1600°C. Experiments were conducted in a sealed vertical tube furnace and the required oxygen partial pressure in the furnace tube was maintained by controlling the flow rates of purified CO and CO2 gas mixtures through the tube. Reaction products were quenched after a reaction time of between 20 and 24 hours, depending on temperature, and the phase compositions were analysed by microprobe. The experimental study revealed that chromium oxide partitions very strongly into the spinel phase relative to the liquid phase, especially at lower temperatures, and higher oxygen partial pressures and bulk chromium oxide contents. The solubility of chromium oxide in the liquid phase was found to increase with increasing temperature and decreasing oxygen partial pressure. An increase in bulk chromium oxide contents of 1 wt%, under otherwise constant conditions, resulted in an increase in slag liquidus temperature of approximately 100°C over the range of temperatures investigated. At 1500°C and bulk chromium oxide contents of 3.7 and 6.4 wt% a reduction in oxygen partial pressure from 1.1x10-5 to 1.1x10-7 atm resulted in increases in soluble chromium oxide of 0.9 and 2.0 wt%, respectively. A further decrease in oxygen partial pressure to 6.7x10-9 atm resulted in increases in soluble chromium oxide of 2.8 and 4.7 wt%, respectively. Experimental results were compared to values predicted by the multi-phase equilibrium (MPE) model developed by CSIRO, and found to agree well. Slag basicity was not varied experimentally and therefore the model was used to predict its effect on the solubility of chromium oxide in the liquid phase and the stability of crystalline phases. At constant temperature, an increase in basicity resulted in a decrease in the solubility of chromium oxide in the liquid phase as well as stabilisation of the spinel phase. It was concluded that practicable combinations of one or more of four main factors, namely increased operating temperature and decreased bulk chromium oxide content, slag basicity and oxygen partial pressure, should be applied and evaluated in a plantenvironment to optimise furnace operation. The MPE model would be a valuable tool in predicting the outcomes of such investigations. / AFRIKAANSE OPSOMMING: Hierdie tesis detaileer die resultate wat verkry is uit ‘n eksperimentele studie uitgevoer om die effek van bedryfstemperatuur, die parsiële druk van suurstof, die algehele chroomoksied inhoud en die algehele FeOx/MgO verhouding op die gedrag van chroomoksied in smelter slakke in die platinum industrie te bestudeer. Twee PGM-bevattende ertslae in die Bosveldkompleks in Suid Afrika, die Merensky en UG2 riwwe, word huidiglik gemyn vir die ekstraksie van basismetale en platinumgroep metale (PGM). Die Merensky rif is ‘n piroksenitiese laag terwyl die UG2 rif ‘n platinumbevattende chromitiet laag is. As gevolg van ‘n geleidelike afname in reserwes van Merensky erts beweeg platinumprodusente al meer na die verwerking van groter hoeveelhede UG2 erts. Die tegniese probleme wat gepaard gaan met die smelting van konsentrate met hoë chroomoksied inhoud kan ‘n rede tot kommer wees. Die vorming van chromiet spinelle in die slak- en matfases verhoog likuidus temperature en viskositeite en bemoeilik die tap van hierdie fases uit oonde. Die opbou van soliede fases verlaag ook die effektiewe oondvolume. Uit die literatuurstudie is gevind dat gepubliseerde studies waarin slak samestellings en parsiële suurstofdrukke wat betrekking het op die platinumindustrie bespreek is, baie beperk is. Dit is gevind dat relevante navorsing gedoen is met aansienlik laer algehele chroom- en ysteroksied konsentrasies. Gevolglik het dit duidelik geword dat ‘n behoefte bestaan vir inligting oor die gedrag van chroomoksied in oonde en die effekte daarvan op fasechemie en –stabiliteit in smelter slakke. Daar is besluit om eksperimente uit te voer deur die gebruik van ses sintetiese slakke met algehele FeOx/MgO verhoudings tussen 0.6 en 1.9 en algehele chroomoksied konsentrasies tussen 1.2 en 7.0 % (op ‘n massabasis). Temperature van 1400, 1500 en 1600°C en suurstof parsiële drukke tussen 6.8x10-10 atm by 1400°C en 8.3x10-5 atm by 1600°C is ondersoek. Eksperimente is uitgevoer in ‘n geseëlde vertikale buisoond en dievereiste suurstofdruk in die oond is gehandhaaf deur beheer van die vloeitempos van gesuiwerde CO en CO2 gas deur die oond. Reaksieprodukte is in water geblus na ‘n reaksietyd van tussen 20 en 24 ure, afhangende van die reaksietemperatuur. Fasesamestellings is bepaal deur mikrosonde analises. Die eksperimentele studie het bewys dat chroomoksied baie sterk in die spinelfase konsentreer relatief tot die vloeistoffase, veral by laer temperature, suurstofdrukke en algehele chroomoksied konsentrasies. Dit is gevind dat die oplosbaarheid van chroomoksied in die vloeistoffase toeneem met toenemende temperatuur en afnemende suurstofdruk. ‘n Toename in die algehele chroomoksied konsentrasie van 1 massa%, onder andersins onveranderde toestande, het ‘n toename van ongeveer 100°C in likuidus temperature veroorsaak tussen 1400 en 1600°C. By 1500°C en algehele chroomoksied konsentrasies van 3.7 en 6.4 massa%, het ‘n verlaging in suurstofdruk vanaf 1.1x10-5 tot 1.1x10-7 atm respektiewelike toenames in die chroomoksied oplosbaarheid van 0.9 en 2.0 massa% veroorsaak. ‘n Verdere verlaging in suurstofdruk tot 6.7x10-9 atm het respektiewelike toenames in chroomoksied oplosbaarheid van 2.8 en 4.7 massa% veroorsaak. Eksperimentele resultate is vergelyk met waardes wat voorspel is deur die multifase ewewigsmodel (MPE), ontwikkel deur CSIRO, en goeie ooreenstemming is gevind. Verskillende slak basisiteite is nie eksperimenteel ondersoek nie en daarom is die model gebruik om die effek daarvan op die oplosbaarheid van chroomoksied in die vloeistoffase asook die stabiliteit van kristallyne fases te bepaal. By konstante temperatuur het ‘n toename in slak basisiteit ‘n afname in chroomoksied oplosbaarheid veroorsaak en die spinelfase gestabiliseer. Die aanbeveling is gemaak dat ‘n kombinasie van een of meer van vier hooffaktore, naamlik hoër bedryfstemperature en laer algehele chroomoksied konsentrasies, slak basisiteit en suurstofdruk, in die praktyk toegepas en geëvalueer moet word om sodoendeoptimum bedryfkondisies te bepaal. Die multifase ewewigsmodel is ‘n nuttige instrument wat gebruik kan word om die resultate van sulke ondersoeke te voorspel.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: This thesis investigates the control of calcium and magnesium in a base metal sulphate leach solution containing nickel and cobalt. The presence of calcium and magnesium in the hydrometallurgical processing of base metals, result in a number of difficulties. These problems range from the contamination of the final product, to high energy consumption and large bleed streams during electrowinning. Calcium poses a greater problem in sulphate solutions due to the low solubility of its sulphate salts. No conventional method exists for the control of calcium and magnesium. As part of this study a review of possible control methods was conducted, which is listed within. From this list the precipitation of fluorides was selected for further investigation. The results showed that it is possible to control calcium and magnesium through the precipitation of their respective fluorides, without the co-precipitation of nickel and cobalt. For 10% stoichiometric excess of fluoride 96.5% calcium and 98.5% magnesium were removed during batch experiments. It is known that mixing and hydrodynamics plays an important role on the characteristics of the formed precipitate, making these processes inherently difficult to scale-up. To evaluate these effects on a continuous process, the three-zone model proposed by Gösele and Kind (1991) was used. A precipitate with consistent characteristics was produced while varying the mixing on the macro, meso and micro scale. Additionally, methods were investigated for the removal or possible recycling of the unreacted fluoride, for which activated alumina was identified. It was observed that activated alumina could adsorb fluoride to low levels in the presence of the base metal solution, after which it could be regenerated again. The activated alumina (AA) had a capacity of 8.65 gF/lAA at a 10 mg/l fluoride breakthrough level during column tests. Based on the experimental results a conceptual process was devised whereby only a portion of the leach stream is subjected to the fluoride precipitation process, after which it is returned to lower the overall calcium and magnesium concentrations. This method would reduce the effect of the observed dominance of magnesium precipitation, in processes where the maximum removal of both elements is not required. The fluoride precipitation process consisted of three steps being precipitation, solid-liquid separation and adsorption of the unreacted fluoride. Sufficient information is provided on the process for a cost estimation to be carried out. Should this found to be feasible, a continuation of the project is recommended. Different reactor configurations could be evaluated for precipitation. The scaling observed during the continuous experiments should also be investigated to minimise its effect. The investigation of activated alumina was only a secondary project and more work is required on optimisation, particularly for the desorption cycle to enable the recycling of the unreacted fluoride. / AFRIKAANSE OPSOMMING: Saamgestel in hierdie tesis is 'n studie van die beheer van kalsium en magnesium in 'n basismetaal-sulfaatoplossing, bevattende nikkel en kobalt. Die teenwoordigheid van kalsium en magnesium in dié oplossings veroorsaak 'n verskeidenheid van probleme, wat wissel van produkkwaliteit verlaging tot hoë energieverbruik en groot bloei strome tydens platering. 'n Groter probleem word ondervind met kalsium as gevolg van die lae oplosbaarheid van sy sulfaatsoute. Geen konvensionele metode kon gevind word vir die beheer van kalsium en magnesium gedurende die oorsig van moontlike metodes nie. Hierdie moontlike metodes is geïdentifiseer en kortliks bespreek in die tesis. Van die moontlike metodes is die presipitasie van fluoried-soute gekies vir verdere eksperimentele ondersoek. Die ondersoek het getoon dat dit moontlik is om kalsium en magnesium te beheer deur die presipitasie van fluoriede sonder om die basismetale saam te presipiteer. Vir 'n 10% oormaat fluoried toevoeging is 96.5% van die kalsium en 98.5% van die magnesium gepresipiteer gedurende die enkelladingstoetse. Dit is bekend dat vermenging en hidrodinamika 'n groot rol speel in die kwaliteit van die presipitaat wat gevorm word. Dit bemoeilik die opskalering van presipitasie prosesse. Vir die ondersoek oor die invloed van vermenging op 'n kontinu proses is die drie-sel model van Gösele en Kind (1991) gebruik. Dit is gevind dat die karakter van presipitaat relatief konstant gebly het vir variasies van vermenging op die makro, meso en mikro skaal, wat opskaling behoort te vergemaklik. Addisioneel is die verwydering of moontlike hersirkulasie van die ongereageerde fluoried ondersoek, en geaktiveerde alumina is geïdentifiseer as 'n moontlike adsorbeermiddel. 'n Eksperimentele ondersoek het getoon dat geaktiveerde alumina fluoried tot lae vlakke kan adsorbeer in die teenwoordigheid van die basismetale, waarna dit weer geregenereer kan word. Die kapasiteit van die geaktiveerde alumina (GA) was bereken as 8.65 gF/lGA by 'n 10 mg/l fluoried vlak gedurende die kolom toetse. 'n Konsep-proses is opgestel na aanleiding van die eksperimentele resultate, waarvolgens slegs 'n gedeelte van die logingstroom na die fluoried presipitasie proses gestuur word, waarna dit weer teruggevoeg word om die algehele kalsium en magnesium konsentrasie te verlaag. Dié metode sal voorkom dat magnesium presipitasie domineer vir 'n toepassing waar slegs 'n gedeelte van kalsium en magnesium verwyder word. Die fluoried presipitasie proses behels drie stappe waarvolgens die fluoriede eers gepresipiteer word, waarna dit geskei word, en dan die ongereageerde fluoried geadsorbeer word. Genoeg inligting is versamel sodat 'n kosteberaming van die proses gedoen kan word. As die koste van die proses aanvaarbaar is, word dit voorgestel dat die ondersoek voortgesit word. Verskillende reaktor konfigurasies kan vir die presipitasie stap getoets word en daar moet ook ondersoek ingestel word hoe om die korslaag wat gedurende die kontinu eksperimente geobserveer is, te verminder. Die ondersoek van geaktiveerde alumina was ondergeskik in die projek en nog werk sal gedoen moet word om dit te optimiseer, spesifiek gedurende die desorpsie siklus vir die hersirkulasie van die ongereageerde fluoried.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: The pneumatic injection of reagent powder into molten iron has become the preferred way to carry out iron and steel desulphurisation. It is therefore essential to not only understand the thermodynamic implications, but also the kinetic principles that govern the desulphurisation process. Key variables that influence the kinetics of the procedure are the condition and composition of the top slag and the melt as well as the injection conditions. Notable injection parameters include reagent flowrate, injection-lance depth and carrier gas flowrate. Owing to sampling restrictions, the subsequent data from Saldanha Steel®, South Africa does not provide adequate insight into the kinetic behaviour of the desulphurisation process and it was therefore the focus of this research to provide an improved quantitive comprehension of the calcium carbide injection procedure at Saldanha Steel. For this purpose a one-dimensional quasi-steady state model for momentum, heat- and mass transfer in rising gas-liquid-powder plumes has been developed for conditions relevant to the Saldanha Steel refining process. Combined with a model predicting the contribution of the topslag to the process, the overall rate of desulphurisation as a function of time can be determined, thus affording the ability to quantitatively explore and analyse the influence of the afore-mentioned injection parameters, as well as the nature of both the topslag and the melt, on the kinetics of the desulphurisation process. Sensitivity analyses concluded that individual increases in the calcium carbide flowrate, the depth of injection and the amount of carry-over slag will result in a reduction in the injection time, while a decrease in the reagent particle diameter and the initial mass of iron in the ladle will have the same effect. Molten iron temperature losses brought about by prolonged injection needs to be electrically recovered within a steelmaking furnace at a high cost. Owing to the high cost of the desulphurising agent, any reduction in the required injection time, while still maintaining product specifications, will therefore result in diminishing overall production costs. Although all the results contained in this study is of particular interest to the Saldanha Steel scenario, it also provides invaluable information and insights into the important variables and parameters playing a role in injection desulphurisation processes in general, along with the influence that changing conditions can have on the end result of such a procedure. / AFRIKAANSE OPSOMMING: Die pneumatiese inspuiting van reagentpoeier is die populêrste ontswawelingsmetode in die yster- en staal bedryf. Dit is dus van groot belang dat die gepaardgaande termodinamiese en kinetiese beginsels betrokke by die ontswawelingsreaksies baie goed verstaan word. Die kondisie en samestelling van die bo-slak en die vloeibare yster, asook die inspuitingkondisies is twee van die belangrikste veranderlikes wat die kinetika van die ontswawelingsproses beïnvloed. Beperkte monsternemingsgeleenthede het veroorsaak dat die relevante data, soos voorsien deur Saldanha Staal®, nie die nodige kinetiese insig in verband met die ontswawelingreaksie weergee nie. Dit is dus die doel van hierdie werkstuk om ‘n verbeterde kwantitatiewe begrip van die ontswawelingsproses by Saldanha Staal daar te stel. Vir hierdie doeleinde is ‘n een-dimensionele, kwasi-gestadigde toestand model vir stygende gas-vloeistof pluime ontwikkel. Die model inkorporeer momentum-, hitte- en massaoordragsprinsiepe en is verteenwoordigend van die ontswawelingsproses by Saldanha Staal. ‘n Tweede model simuleer die bydrae wat die bo-slak tot die algehele ontswawelingsproses maak en saam gee hierdie twee modelle die algehele ontswawelingstempo weer as ‘n funksie van tyd. Die modelle word ook gebruik om die invloed van die bogenoemde inspuitingsveranderlikes op die proses te ondersoek. Deeglike sensitiwiteitsanalise het gewys dat ‘n verhoging in die kalsium karbied vloeitempo, asook die inspuitingsdiepte van die lans en die hoeveelheid slak wat vanaf die boogoond na die ontswawelingseenheid oorgedra word, ‘n vermindering in die vereisde inspuitingstyd te weeg bring. Verkleining in die kalsium kardied partikels se gemiddelde diameter en vermindering van die hoeveelheid yster in die torpedokarre aan die begin van die proses, het dieselfde uitwerking op die vereisde inspuitingstyd. Geweldig baie geld moet aan elektrisiteit spandeer word om die temperatuur wat verlore gaan as gevolg van onnodige lang inspuitingstye, in die staalmaakoonde te herwin. Gekombineerd met die feit dat die kalsium karbied reagent baie duur is, beteken dit dat reduksies in die vereisde ontswaweling inspuitingstyd groot besparings te weeg kan bring. Alhoewel die saamgevatte resultate van spesifieke belang is vir die Saldanha Staal proses, verskaf hierdie studie waardevolle informasie oor die belangrikheid van verskeie veranderlikes, asook die rol wat veranderende toestande op die eindresultate van die ontswawelingproses kan hê.

Dissertation (PhD)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: Silver/Zinc batteries are extensively used for applications where a power source with a high energy density is needed. A reactor was developed to manufacture the zinc electrodes for these batteries. The parameters for the control of this reactor were identified. The interaction and relationships between the set of parameters were established and quantified. The novel design of the reactor includes features such as continuous replenishment of the constituting chemicals, parallel flow of the plating fluid, inert anodes and a serial electrical layout. The controlling parameters consist of both chemical and physical parameters. They are - zincate concentration, hydroxyl concentration, density, temperature, flow rate of the plating fluid, and current density. The unique solubility characteristics of zinc oxide in potassium hydroxide lend themselves to a novel approach for the determination of the zincate concentration in· the plating solution. A set of conditions for the manufacturing of electrodes, with the required characteristics, was identified. The requirements are densely packed zinc dendrite morphology with high porosity and mechanical stability. The designed reactor complies with the industries safety requirements. The batteries were successfully applied in their operational domain (torpedoes) after this extensive research project was completed. / AFRIKAANSE OPSOMMING: Silwer/Sink batterye word intensief aangewend in gebruike waar 'n kragbron met hoë energie digtheid benodig word. 'n Reaktor is ontwikkel wat die sink elektrodes van die tipes batterye vervaardig. Die veranderlikes vir die beheer van die reaktor is bepaal. Die interaksie en verwantskap tussen die stel veranderlikes is verkry en gekwantifiseer. Die unieke ontwerp van die reaktor het die volgende karakteristieke, nl. aaneenlopende vervanging van die chemikalië verbruik, parallelle vloei van die plateringsvloeistof, inerte anodes en elektriese uitleg in serie. Die chemiese en fisiese veranderlikes is beide beherende veranderlikes. Hierdie fisiese en chemiese veranderlikes wat die proses beheer is die sinkaat konsentrasie, hidroksiel konsentrasie, digtheid, temperatuur, vloeitempo van die plateringsvloeistof en die stroomdigtheid. Die unieke oplossingskarakteristieke van sinkoksied in kalsium· hidroksied leen hom tot die unieke proses om die sinkaat konsentrasie in die plateringsvloeistof te bepaal. 'n Steloperasionele kondisies vir die vervaardiging van die elektrodes met die spesifieke spesifikasies soos verlang, is geïdentifiseer. Hierdie kondisies is die diggepakte sinkdendriet morfologie met 'n hoë porositeit en meganiese stabiliteit. Die ontwerpte reaktor voldoen aan die industriële veiligheidsvereistes en die ergonomiese maatstawwe. Die batterye is suksesvol gebruik in hulle industriële aanwending as gevolg van die suksesvolle afhandeling van die stuk navorsing.

Thesis (PhD)--University of Stellenbosch, 2002. / ENGLISH ABSTRACT: The availability of vast deposits of high-grade ore bodies are rapidly becoming something of the past in the modern mining and metallurgical scenario. Apart from the lower grade content of these ore bodies, complex mineralogy are an even greater obstacle in the recovery of valuable metal species. The development of new technology to deal with these type of ore bodies is therefore critical and worth investigating, as the world's easily exploitable high grade ore deposits are decreasing. Valuable species can be recovered from sparingly soluble solids, which slightly dissociate to give traces of the valuable ions in solution, with the use of ion-exchange resins in a slurry mixture. A dissociation equilibrium exists between the dissolved ions in solution and the solid ore body. Jf the dissolved ions are removed from the solution by ion-exchange, the solid / liquid dissociation equilibrium is continually displaced. According to Le Chatelier's principle further dissolution of the sparingly soluble solid is required to restore the equilibrium concentration of the valuable species in solution. It is possible to recover valuable metal species from metal precipitates, such as metal sulphides, by contacting a slurry of the precipitate with a suitable ion-exchange resin. The resulting ion exchange reaction between the valuable metal species and counter ions creates electrolyte solutions that may facilitate the further dissolution of the metal precipitate. These counter ion electrolyte solutions may easily become significantly concentrated. This occurs in the event of a Resin-in-Leach (RIL) mixture that results in a continuous ion-exchange reaction taking place due to the continually changing electrolyte composition of the mixture, which significantly changes the activities and hence the solubility of the valuable metal species in solution. Complete dissolution and liberation of the metal precipitate can often be achieved provided that a sufficient amount of a suitable high capacity ion-exchange resin is used in a properly engineered Resin-in- Leach (RIL) circuit. The simultaneous dissolution and adsorption of various base metal precipitates were tested. Various interactions that take place in the slurry at molecular level as well as the effects of various variables on the "adsorption by dissolution process" are discussed through the development of fundamental thermodynamic models. These thermodynamic mathematical models are developed for the three phase system that exists in a Resin-in- Leach mixture, i.e. the solid ore body, the electrolyte solution and the ion-exchange resin, and can be used for possible other applications such as the recovery of rare earths from low grade ores in the minerals processing industry. A typical example of an industrial process for the recovery of rare earth species is the percolation leaching of rare earths from low-grade kaolinitic ores, which continually shifts the solid / liquid dissociation equilibrium condition. The rare earth content of these ores is usually between 0.05%and 0.3 %, which is very low by any modern industrial extraction and refining standards. / AFRIKAANSE OPSOMMING: Die beskikbaarheid van ryk mineraal ertsneerslae is spoedig besig om iets van die verlede te word in die huidige mineraalontginning en metallurgiese veredelings industrie. Afgesien van die lae graad van die huidige mineraal ertsneerslae, blyk die komplekse mineralogiese samestelling van hierdie neerslae In veel groter struikelblok te wees in die herwinning en veredeling van die edelmetale teenwoordig in hierdie ertse. Die ontwikkeling van nuwe veredelings en ekstratiewe tegnologie vir die herwinning van edel metale, vanuit hierdie lae graad mineraal ertsneerslae, word dus benodig wat verdere navorsing in hierdie gebied regverdig. Dit is wel moontlik om metaal spesies afkomstig van ertse met 'n baie lae oplosbaarheid in waterige oplossings te herwin met ioon-uitruilings harse vanweë die feit dat die metaal spesies wel teen baie lae konsentrasies in die waterige oplossings teenwoordig is. Die metaal spesies los op in die waterige oplossings volgens hulle karakteristieke oplossings termodinamika. Indien die opgeloste metaal spesies vanuit die waterige oplossing verwyder word, sal die vaste stof / vloeistof ewewigs balans weer herstel word deurdat die vaste stof verder saloplos as gevolg van Le Chatelier se beginsel. Dit is dus moontlik om metaalagtige spesies, soos metaal sulfiedes, te herwin deur 'n waterige oplossing wat die metaal erts bevat in kontak te bring met 'n ioon-uitruilings hars. Die daaropvolgende ioon-uitruilings reaksie tussen die metaalagtige spesies en die spesies teenswoordig op die ioon-uitruilingshars het tot gevolg dat die elektrolitiese samestelling van die waterige oplossing verander. Dit is die gevolg van die migrasie van spesies aanvanklik teenswoordig op die hars wat in die waterige fase eindig. Die veranderende samestelling van die waterige oplossing mag verder tot gevolg hê dat die oplosbaarheid van die metaalagtige spesie verder verhoog mag word. Die gevolg van bogenoemde reaksies is dat die waterige oplossing ionies sterk gekonsentreerd kan word soos meer en meer spesies aanvanklik teenswoordig op die hars migreer na die oplossing. Die ioniese verandering van die waterige oplossing van 'n suiwer waterige fase tot 'n ionies sterk gekonsentreerde oplossing vind plaas tydens die oplos van erste in 'n hars-inpulp (HIP) oplossing. Die nuwe ioniese aktiwiteit in die oplossing kan die oplosbaarheid van die vaste stof drasties verander. Die volledige oplossing van 'n bepaalde kwantiteit van die vaste stof kan bereik word deur die genoegsame toevoeging van 'n geskikte hars tot die waterige oplossing wat die vaste stofbevat. Die gelyktydige oplossing en absorpsie van die metaalagtige vastestowwe vanuit waterige oplossings met behulp van ioon-uitruilings harse is eksperimenteel getoets vir die doeleindes van hierdie werkstuk. Verskeie interaksies wat op molekulêre vlak in die pulp plaasvind asook die adsorpsie proses van die spesies vanuit die waterige oplossing op die harse word bespreek en gemodelleer. Wiskundige modelle wat die termodinamika van die verskillende fases wiskundig verteenwoordig is ontwikkel vanaf bestaande termodinamiese beginsels. Die drie verskillende fases wat in ag geneem is, is die waterige fase met opgeloste metaal spesies, die vaste stof fase wat as die erts in bostaande paragrawe beskryf is en die ioon-uitruilings fase wat 'n komplekse vastestof en water fase gekombineerd is. Die doel van die werkstuk is om die basis te skep vir die ontwikkeling van modelle wat gebruik kan word om die herwinning van skaars-aarde mertale vanuit lae oplosbare erstse te modelleer en beskryf. 'n Tipiese industriële voorbeeld is die herwinning van skaars-aarde metale van lae-graadse kalkagtige erste deur gebruik te maak van perkolerende logings reaksies wat geduring die vastestof / vloeistof ewewig versteur. Die konsentrasie van die skaars-aarde metale in hierde erts gesteentes kan wissel vanaf so laag as 0.05% tot 0.3% per massa. Hierdie konsentrasies is uiters laag gesien vanuit enige industriële herwinnings proses oogpunt.

Thesis (PhD)--Stellenbosch University, 1992. / ENGLISH ABSTRACT: The dissolution of gold from refractory ores is a complex kinetic problem involving a number of chemical, mass transport and mineralogical factors. In most Witwatersrand ores in South Africa more than 97 % of the gold is dissolved in cyanide medium after a residence time of about 16 hours in pachuca tanks. This high percentage may be the reason why so little fundamental research has been done into the mechanism and kinetics of the leaching process. With the increasingly lower grades of ore mined, the introduction of backfill mining, and the reduction of profit margins, it has become imperative to increase the efficiency of gold dissolution. The effects of the chemistry and particle size on the dissolution of gold in each sample of ore were studied in detail. The emphasis in this study is on the effect of the leaching behaviour of various ore constituents on the rate of gold dissolution. Interferences with the leaching of gold in contact with other minerals or metals could be attributed to the galvanic interaction (electrical conductivity) between the gold and the mineral and to the formation of a surface film on the gold surface. Sulphide minerals and their oxidation products cause the largest decrease in gold dissolution rate. Galena enhances the rate of gold dissolution owing to dissolved Pb(II)-ions. Gold in contact with conductive minerals passivates as a result of the enhanced magnitude of the cathodic cu1Tent. In all experiments the rotating disc of gold passivated so that the rate of dissolution was much slower than that predicted by a mass-transport limiting model. The various films that form on the surface of the gold and associated minerals, as well as the galvanic interaction, depend largely on the pretreatment of the ore. Pre-elimination of host minerals from the gold bearing ore increases the dissolution rate of gold, and explains the kinetics of reaction on the gold surface to a large extent. The selective destruction of the various minerals with oxidative acid leaches destroys and/or decomposes certain minerals which may form films on the gold surface by precipitation. The chemical composition of these films and precipitates depends on the mineralogy of the sample. These films may be oxides, sulphides, carbonates and cyanide complexes. The complexes can be destroyed, depending on the nature of the film, by interstage dilute acid and/or cyanide washes in an agitated vessel. The destruction of the films exposes the gold surface for cyanidation. A simple distribution function similar to the King liberation model is proposed and tested to describe the dissolution step in the multi-step leaching mechanism. For the King model, good agreement is shown with experimental results. For the liberation results obtained by leaching in this study, the trend is co1Tect, but calibration is required for a close fit. A potentially important use for the liberation model by leaching is to predict the leachable or free gold in an ore from the free gold in the complete sample. This approach for studying the leaching behaviour of different gold bearing minerals has provided reasons why some ores leach better than others. / AFRIKAANSE OPSOMMING: Die loging van goud vanuit weerbarstige ertse is 'n komplekse kinetiese probleem wat verskeie faktore soos massa-oordrag, chemiese aspekte en mineralogiese ingeweefdheid insluit. Goud ekstraksies so hoog as 97 % in sianied oplossings in Pachuca reaktore na ongeveer 16 uur logingstyd word behaal in die meeste Witwatersrand ertse in Suid-Afrika. Hierdie hoë ekstraksies mag dalk die rede wees vir die min fundamentele navorsing oor die ekstraksie van goud vanuit minerale in 'n spesifieke erts. Die dalende erts grade, die verlaging van winsgrense en die terugplaas van geloogde erts in die myn noodsaak verbeterde goud ekstraksie. Verkeie faktore nl., chemie, diffusie, partikelgrootte en oplosbaarheid van goud in elke monster erts is in hierdie studie ondersoek. Die sentrale tema was om die logingsgedrag van goud vanuit verskeie minerale in 'n erts te bepaal. Galvaniese interaksie (hou verband met elektriese geleidingsvermoeë) en film vorming is die belangrikste faktore wat die loging van goud in kontak met minerale nadelig beïnvloed. Sulfied minerale en hul oksidasie produkte speel die grootste rol in die verlaging van die tempo van goudloging. Galena verhoog die tempo van goudloging as gevolg van die Pb (II)- ione in oplossing. Goud in kontak met geleidende minerale passiveer as gevolg van die verhoogde katodiese stroomdigtheid. In alle eksperimente met die roterende skyf (goudskyf) apparaat, passiveer die goudskyf in so 'n mate dat die logingstempo baie stadiger is as wat voorspel word met die massa-oordrags model. Die onderskeie films wat vorm op die goud- en geassosieerde minerale se oppervlaktes, asook die galvaniese interaksies, is 'n funksie van die voorafbehandeling van die erts. Die selektiewe eliminering van minerale vanuit 'n gouddraende erts verhoog die tempo van goudloging drasties en dit beskryf die kinetika van goudloging op die goudoppervlak in 'n groot mate. Die selektiewe eliminering van minerale deur gebruik te maak van oksiderende suurlogings, vernietig sekere van die minerale wat films op die goudoppervlakte kan veroorsaak deur middel van presipitasie. Die chemiese samestelling van hierdie films hang af van die mineralogie van die monster. Dit bestaan meestal uit oksiedes, sulfiedes, karbonate en sianiedkomplekse en hulle kan vernietig word deur middel van inter-stadia verdunde suur-en/of sianied wasse. 'n Eenvoudige distribusiefunksie, soortgelyk aan die King bevrydingsmodel word voorgestel en eksperimenteel getoets om die logingstap in die multi-stadia logingsmeganisme te beskryf. Vir bevryding deur loging, is die neiging van King se model korrek, maar kalibrasie word benodig vir goeie passing. 'n Potensiele gebruik van die aangepaste model is om vrye of loogbare goud in 'n spesi fieke partikel grootte fraksie van 'n erts te voorspel as die vry goud in die totale fraksie bekend is. Die resultate uit hierdie studie kan gebruik word om die logingsgedrag van gouddraende minerale te voorspel en te beskryf, en dus veduidelik hoekom goud uit sekere ertse beter loog as uit ander.

Thesis (MScIng)--University of Stellenbosch, 2003. / ENGLISH ABSTRACT: Physical systems encountered in process engineering are invariably ill-defined, multivariate, and exhibit complex nonlinear dynamical behaviour. The increasing demands for better process efficiency and high product quality have led to the development and implementation of advanced control strategies in process plants. These modern control strategies are based on the use of a mathematical model defined for the process. Traditionally, linear models have been used to approximate the dynamics of processes whereas most processes are governed by nonlinear mechanisms. Since linear systems theory is well-established whereas nonlinear systems theory is not, recent developments in nonlinear dynamical systems theory present opportunities for improved approaches in modelling these process systems. It is now known that a nonlinear description of a process can be obtained from using time-delayed copies reconstructed from measurements taken from the process. Due to low signal to noise ratios associated with measured data it is logical to exploit redundant information in multivariate time signals taken from the systems in reconstructing the underlying dynamics. This study investigated the extension of univariate nonlinear time series analysis to the situation where multivariate measurements are available. Using simulated data from a coupled continuously stirred tank reactor and measured data from a flotation process system, the comparative advantages of using multivariate and univariate state space reconstructions were investigated. With respect to detection of nonlinearity multivariate surrogate analysis were found to give potentially robust results because of preservation of cross-correlations among components in the surrogate data. Multivariate local linear models showed a deterministic structure in both small and large neighbourhood sizes whereas for scalar embeddings determinism was defined only in smaller neighbourhood sizes. Non-uniform multivariate embeddings gave local linear models that resembled models from a trivial reconstruction of the original state space variables. With regard to global nonlinear modelling, multivariate embeddings gave models with better predictability irrespective of the model class used. Further improvements in the performance of models were obtained for multivariate non-uniform embeddings. A relatively new statistical learning algorithm, the least-squares support vector machine (LSSVM), was evaluated using multilayer perceptrons (MLP) as a benchmark in modelling nonlinear time series using simulated and plant data. It was observed that in the absence of autocorrelations in the variables and sparse data LSSVMs performed better than MLPs. Simulation of trained models gave consistent results for the LSSVMs, which was not the case for MLPs. However, the computational costs incurred in training the LSSVM model was significantly higher than for MLPs. LSSVMs were found to be insensitive to dimensionality reduction methods whereas the performance of MLPs degraded with increasing complexity of the dimension reduction method. No relative merits were found for using complex subspace dimension reduction methods for the data used. No general conclusions could be drawn with respect to the relative superiority of one class of models method over the other. Spatiotemporal structures are routinely observed in many chemical systems, such as reactive-diffusion and other pattern forming systems. We investigated the modelling of spatiotemporal time series using the coupled logistic map lattice as a case study. It was found that including both spatial and temporal information improved the performance of the fitted models. However, the superiority of spatiotemporal embeddings over individual time series was found to be defined for certain choices of the spatial and temporal embedding parameters. / AFRIKAANSE OPSOMMING: Fisiese stelsels wat in prosesingenieurswese voorkom is dikwels nie goed gedefinieer nie, multiveranderlik en vertoon komplekse nie-lineˆere gedrag. Toenemende vereistes vir ho¨e prosesdoeltreffendheid en produkgehalte het gelei tot die ontwikkeling en implementering van gevorderde beheerstrategie¨e vir prosesaanlegte. Hierdie morderne beheerstrategie¨e is gebaseer op die gebruik van wiskundige prosesmodelle. Lineˆere modelle word gewoonlik ontwikkel, al is die onderliggende prosesmeganismes in die algemeen nie-lineˆere, aangesien lineˆere stetselteorie goed gevestig is, en nie-line¨ere stelselteorie nie. Onlangse verwikkelinge in die teorie van nie-lineˆeredinamiese stelsels bied egter geleenthede vir verbeterde modellering van prosesstelsels. Dit is bekend dat ‘n nie-lineˆere beskrywing van ‘n progses verkry kan word deur tydvertraagde kopie¨e van metings van die prosesse te rekonstrueer. Met die lae seintot- geraasverhoudings wat met gemete data geassosieer word, is dit logies om die oortollige informasie in meerveranderlike seine te benut tydens die rekonstruksie van die onderliggende prosesdinamika. In die tesis is die uitbreiding van enkel-veranderlike nie-lineˆere tydreeksontleding na meer-veranderlike stelsels ondersoek. Met data van twee aaneengeskakelde gesimuleerde geroerde tenkreaktore en werklike data van ‘n flottasieproses, is die meriete van enkel- en meerveranderlike rekonstruksies van toestandruimtes ondersoek. Meerveranderlike surrogaatdata-ontleding het nie-lineariteite in die data op ‘n meer robuuste wyse ge¨ıdentifiseer, a.g.v. die behoud van kruis-korrelasies in die komponente van die data. Meerveranderlike lokale lineˆere modelle het ‘n deterministiese struktuur in beide klein en groot naasliggende omgewings ge¨ıdentifiseer, terwyl enkelveranderlike metodes dit slegs vir klein naasliggende omgewings kon doen. Nie-uniforme meerveranderlike inbeddings het lokale lineˆere modelle gegenereer wat soos globale modelle afkomstig van triviale rekonstruksies van die data gelyk het. M.b.t globale nie-lineˆere modellering, het meerveranderlike inbedding deurgaans beter modelle opgelewer. Verdere verbetering in die prestasie van modelle kon verkry word d.m.v. meerveranderlike nie-uniforme inbedding. ‘n Relatief nuwe statistiese algoritme, die kleinste-kwadrate-steunvektormasjien (KKSVM) is ge¨evalueer teenoor multilaag-perseptrons (MLP) as ‘n standaard vir die modellering van nie-lineˆere tydreekse, deur gebruik te maak van gesimuleerde en werklike aanlegdata. Daar is gevind dat die KKSVM beter presteer het as die MLPs wanneer die opeenvolgende waarnemings swak gekorreleer en min was relatief tot die aantal veranderlikes. Die KKSVMs het beduidend langer geneem as die MLPs om te ontwikkel. Hulle was ook minder sensitief vir die metodes wat gevolg is om die dimensionaliteit van die data te verlaag, anders as die MLPs. Ook is gevind dat meer komplekse metodes tot die verlaging van die dimensionaliteit weinig nut gehad het. Geen algemene gevolgtrekkings kan egter gemaak word m.b.t die verskillende modelle nie. Ruimtelik-temporale strukture word algemeen waargeneem in baie chemiese stelsels, soos reaktiewe diffusie e.a. patroonvormende sisteme. Die modellering van ruimtelik-temporale stelsels is bestudeer aan die hand van ‘n gekoppelde logistiese projeksierooster. Insluiting van beide die ruimtelike en temporale inligting het tot beduidend beter modelle gelei, solank as wat di´e inligting op die regte wyse ontsluit is.

Thesis (MScEng)--University of Stellenbosch, 2002. / ENGLISH ABSTRACT: Metallurgical processes are often high dimensional and non-linear making them difficult to understand, model and control. Whereas the human eye has extensively been used in discerning temporal patterns in historical process data from these processes, the systematic study of such data has only recently come to the forefront. This resulted predominantly from the inadequacy of previously used linear techniques and the computational power required when analysing the non-linear dynamics underlying these systems. Furthermore, owing to the recent progress made with regard to the identification of non-linear systems and the increased availability of computational power, the application of non-linear modelling techniques for the development of neural network models to be used in advanced control systems has become a potential alternative to operator experience. The objective of this study was the development ofa non-linear, dynamic model of an autogenous mill for use in an advanced control system. This was accomplished through system identification, modelling and prediction, and application to control. For system identification, the attractor was reconstructed based on Taken's theorem making use of both the Method Of Delays and singular spectrum analysis. Modelling consisted of the development of multi-layer perceptron neural network, radial basis function neural network, and support vector machine models for the prediction of the power drawn by an autogenous mill. The best model was subsequently selected and validated through its application to control. This was accomplished by means of developing a neurocontroller, which was tested under simulation. Initial inspection of the process data to be modelled indicated that it contained a considerable amount noise. However, using the method of surrogate data, it was found that the time series representing the power drawn by the autogenous mill clearly exhibited deterministic character, making it suitable for predictive modelling. It was subsequently found that, when using the data for attractor reconstruction, a connection existed between the embedding strategy used, the quality of the reconstructed attractor, and the quality of the resulting model. Owing to the high degree of noise in the data it was found that the singular spectrum analysis embeddings resulted in better quality reconstructed attractors that covered a larger part of the state space when compared to the method of delays embeddings; the data embedded using singular spectrum analysis also resulting in the development of better quality models. From a modelling perspective it was found that the multi-layer perceptron neural network models generally performed the best; a multi-layer perceptron neural network model having an appropriately embedded multi-dimensional input space outperforming all the other developed models with regard to free-run prediction success. However, none of the non-linear models performed significantly better than the ARX model with regard to one-step prediction results (based on the R2 statistic); the one-step predictions having a prediction interval of 30 seconds. In general the best model was a multi-layer perceptron neural network model having an input space consisting of the FAG mill power (XI), the FAG mill load (X2), the FAG mill coarse ore feed rate (X3), the FAG mill fine ore feed rate (X4), the FAG mill inlet water flow rate (X7) and the FAG mill discharge flow rates (X9, XIO). Since the accuracy of any neural network model is highly dependent on its training data, a process model diagnostic system was developed to accompany the process model. Linear principal component analysis was used for this purposes and the resulting diagnostic system was successfully used for data validation. One of the models developed during this research was also successfully used for the development of a neurocontroller, proving its possible use in an advanced control system. / AFRIKAANSE OPSOMMING: Metallurgiese prosesse is gewoonlik hoogs dimensioneel en nie-lineêr, wat dit moeilik maak om te verstaan, modelleer, en te beheer. Alhoewel die menslike oog alreeds wyd gebruik word om temporale patrone in historiese proses data te onderskei, het die sistematiese studie van hierdie tipe data eers onlangs na vore gekom. Dit is hoofsaaklik na aanleiding van die onvoldoende resultate wat verkry is deur van voorafgaande lineêre tegnieke gebruik te maak, asook die beperkende berekenings vermoë wat beskikbaar was vir analise van onderliggend nie-lineêre dinamiese stelsels. 'n Verder bydraende faktor is die onlangse vordering wat gemaak is met betrekking tot die identifikasie van nie-lineêre stelsels en die toename in beskikbaarheid van rekenaar-vermoë. Die toepassing van nie-lineêre modellerings tegnieke vir die ontwikkeling van neurale netwerke om gebruik te word in gevorderde beheerstelsels, het 'n potensiële alternatief geword tot operateur ondervinding. Die doelwit van hierdie studie was die ontwikkeling van 'n gevorderde beheerstelsel vir 'n outogene meul gebaseer op 'n nie-lineêre, dinamiese model. Dit is bereik deur middel van stelsel-identifikasie, modellering en voorspelling, en laastens implementering van die beheerstelsel. Vir stelsel-identifikasie is die attraktor van die stelsel bepaal soos gebaseer op Taken se teorema deur gebruik te maak van beide die metode van vertraging en enkelvoudige spektrum analise. Modellering van die stelsel vir die voorspelling van krag-verbruik deur die outogene meul het bestaan uit die ontwikkeling van multilaag-perseptron-neurale netwerke, radiaalbasisfunksie-neurale netwerke, en steunvektor-masjien-modelle. Die beste model is daarna gekies vir validasie deur middel van toepassing vir beheer. Dit is bereik deur 'n neurobeheerder te ontwikkel en te toets deur middel van simulasie. Die aanvanklike inspeksie van proses data wat gebruik sou word vir modellering het egter getoon dat die data 'n aansienlike hoeveelheid geraas bevat. Nietemin, deur die gebruik van 'n surrogaat-data-metode, is dit bevind dat die tyd-reeks wat die krag verbruik van die outogene meul beskryf, duidelik deterministiese karakter toon en dat dit dus wenslik is om 'n nie-lineêre voorspellings-model, soos 'n neurale netwerk te gruik. Gevolglik is gevind dat, wanneer die data vir attraktor hersamestelling gebruik word, 'n verband bestaan tussen die ontvouing-strategie wat gebruik word, die kwaliteit van die gerekonstrueerde attraktor, en die kwaliteit van die daaropvolgende model. As gevolg van die geraas in die data is gevind dat die ontvouing gebaseer op enkelvoudige spektrum analise 'n beter kwaliteit attraktor hersamestelling lewer. So ook is gevind dat 'n groter deel van die toestandruimte gedek word in vergelyking met die metode van vertraging-ontvouing. Deur gebruik te maak van enkelvoudige spektrum-analise, het die dataontvouing ook beter kwaliteit modelle opgelewer. Vanuit 'n modellerings-perspektief is gevind dat die multilaag-perseptron-neurale netwerk-modelle in die algemeen die beste gevaar het. 'n Multilaag-perseptronneurale netwerk met 'n gepaste ontvoude multidimensionele invoer-spasie het die beste gevaar van al die ontwikkelde modelle met betrekking tot vryloopvoorspellings. Geen van die nie-lineêre modelle het egter beduidend (op 'n R2 basis) beter gevaar as die ARX model wanneer daar na die eenstap-voorspellings (oor 'n 30 sekonde interval) gekyk word nie. Die multilaag-perseptron-neurale netwerk met 'n invoer-spasie bestaande uit die meul krag-verbruik (XI), die meullading (X2), die meul growwe-erts voertempo (X3), die meul fyn-erts voertemp ('4), die meul inlaat-water vloeitempo (X7) en die meul uitlaat vloeitempo's (X9, XIO) het in die algemeen die beste gevaar. Aangesien die akkuraatheid van emge neurale netwerk afhanklik is van die data waarmee dit aanvanklik opgestel is, is 'n diagnostiese proses modelontwikkel om die proses-model te vergesel. Lineêre hoofkomponent analise is vir hierdie doel aangewend en die gevolglike diagnostiese stelsel is suksesvol aangewend vir datavalidasie. Een van die modelle ontwikkel gedurende hierdie navorsing is ook suksesvol gebruik vir die ontwikkeling van 'n neurobeheerder wat dien as bewys dat die model goed gebruik kan word in 'n gevorderde beheerstelsel.