Search Documents

By H. W. Glen

The use of a personal style (mainly active voice and first person) makes technical writing more pleasant to read, and helps the writer to express his meaning clearly and to avoid many of the stylistic errors normally encountered in that type of writing.

Jan 1, 1979

By L. von Ketelhodt

During the period October 2003 to June 2004, a containerized optical sorting pilot plant was operated at Kloof Gold Mine on a waste rock dump. In this paper, we describe the test work and its results leading to the decision to operate this pilot plant. This paper presents the operational data and financial figures achieved during this project, in particular, for the month of June 2004 when the plant was run at full capacity on three shifts, 24 hour operation. These results are also extrapolated and presented to show the profitability of such an operation at today?s gold price levels and cost structure.

Jan 1, 2009

By T. Scheldt, M. Lu, A. Myrvang

Abstract: In this paper both continuous and discontinuous modelling are utilised in order to analyze the stability of the Gjøvik Olympic Mountain Hall. The large span cavern (62 m) was built for the 1994 Winter Olympic Games at Lillehammer, and is a unique case study because of the large amount of available input data, through a comprehensive pre-investigation, stress and deformation monitoring program before, during and after the excavation. Phase2, a two-dimensional non-linear finite element program, and the Discontinuous Deformation Analysis (DDA) are used as representative tools for continuous and discontinuous modelling, respectively. The analysis results are compared and discussed taking the reliability of the models, time-consumption and costs into consideration. Some guidelines for the total modelling process, from field measurements to final results, are also presented. Zusammenfassung: In diesem Artikel werden sowohl kontinuierlich als auch diskontinuierlich Modellierung verwendet, um die Stabilität, der in den Berg gebauten Olympiahalle in Gjøvik zu analysieren. Die Olympiahalle wurde 1994 zu den Olympische Spielen in Lillehammer errichtet. Ihr Dach erstreckt sich über eine Breite von zweiundsechzig Meter. Für den Bau wurden umfangreiche Daten durch Voruntersuchungen, und Überwachung von Gebirgsspannungen und Deformationen bevor, während und nach Exkavation eingesammelt. Auf Grund dieser großen Datenmenge stellt die Olympiahalle ein herausragendes Beispiel für den Bau ähnlicher Projekte dar. Für die kontinuierlichen Modellierungen wurde „Phase2“, ein zwei-dimensionales, nicht lineares „endlich Element“ Rechnerprogramm angewandt. Die diskontinuierlichen Modellierungen erstellte man mit Hilfe von der „Diskontinuierliche Deformations Analyse“ (DDA). Die Analysenergebnisse sind mit Rücksicht auf die Zuverlässigkeit der Modelle, Zeitverbrauch und Kosten verglichen und diskutiert worden. Im Artikel werden Richtlinien für den Modellierungsprozess von Dateneinsammlung bis Schlussfolgerung gegeben. Résumé: Dans cet article la modélisation continue et discontinue sont utilisés pour analyser la stabilité de Gjøvik Olympic Mountain Hall. Cette caverne de grande portée (62 m) a été construite pour les Jeux Olympiques d’Hiver de 1994 à Lillehammer, et est une étude de cas unique grâce à la grande quantité de données disponibles, de par un programme de pré-investigation compréhensive, monitorage des contraintes et déformations avant, pendant et après l'excavation. Phase2, un code d’éléments finis non linéaire à deux dimensions, et l’Analyse en Déformations Discontinues (DDA) sont utilisés comme outils représentatifs de la modélisation continue et discontinue, respectivement. Les résultats d'analyses sont comparés et sont discutés, prenant en compte la fiabilité des modèles, la consommation en temps de calcul et les coûts. Quelques indications pour le procédé total de modélisation, des mesures sur place aux résultats finaux, sont aussi présentées.

Jan 1, 2003

By I. Mitov, A. Stoilova, D. Krastev, B. Yordanov

We present three technological scenarios for the recovery of valuable components from gangue, stored in the tailings dam at Kremikovtzi metallurgical plant in Bulgaria, into marketable iron-containing pellets. In the first approach the iron concentrate was recovered through a two-stage flotation process, desliming, and magnetic separation. In the second proposed process, the iron concentrate was subjected to four sequential stages of magnetic separation coupled with selective magnetic flocculation. The third route entails the not very common practice of magnetizing roasting, followed by selective magnetic flocculation, desliming, and magnetic separation. The iron concentrate was pelletized in a laboratoryscale pelletizer. Each technology has been assessed with regard to the mass yield of iron concentrate, the iron recovery. and the iron, lead, and zinc content in order to identify the most effective route.

May 1, 2021

This paper describes the extraction of gold, as the anionic aurocyanide complex, from solution by strong- and weakbase anion-exchange resins. The effects, on the rate of extraction and the loading capacity for gold, of parameters such as pH, temperature, ionic strength, agitation, and the presence of competing anions are reported. The elution and regeneration of strong- and weak-base resins will be examined in a second paper. A third paper in this series will present the results of several smallscale pilot-plant investigations carried out with strong- and weakbase resins In cyanide pulps from various South African gold mines, and will compare the metallurgical performance of resin-In-pulp with that of the carbon-in-pulp process.

Jan 1, 1984

By D. Yu Pleev, V. G. Krivolapov

Mines of Southern Kuzbass working out, basically, coal formations of flat seams. Seams, as a rule, are inclined to self-ignition. Detection at early stages of the centers of endogenous underground fires has great value for safe conducting mountain works. One of methods of detection of an early stage of underground fires in the worked out space is the method of superficial radon survey. In the report the method is de-scribed and examples of superficial radon survey on concrete layers are given.

Jan 1, 2003

By R. C. Nkhoma, K. A. Annan, S. Ngomane

The effects of welding current, electrode force, and welding time in a resistance spot weld were studied to investigate the effectiveness of welded joints between a thin EN10130: DC04 material and a thicker 817M40 part, through analysis of the microstructural and mechanical properties. All welded specimens were subjected to tensile testing at room temperature (25°C) and sub-zero temperature (–46°C) to test the strength of the welded joints. No full button failure was observed at either room temperature or sub-zero temperature after optimization of the weldng parameters. The fusion zone was observed to consist mainly of martensitic phase, due to rapid quenching, while the HAZ was composed of clusters of martensite in a ferrite and bainite matrix. The base 817M40 metal remained fully ferritic after welding. The hardness was found to increase with increasing welding current. An increase in nugget size, indicating good fusion of the weld, was observed with an increase in the welding current.

Oct 1, 2021

By S. T. L. Harrison, A. Kotsiopoulos, M. D. Maluleke, E. Govender-Opitz

In the extraction of base metals from printed circuit boards (PCBs) using bioleaching, the accumulation of metal ion stress in solution has been shown to inhibit the microbial regeneration of ferric iron. This poses a challenge for the development of this technology for e-waste treatment. The paper explores the application of a microbial immobilisation system in a two-stage bioleaching system consisting of a chemical leaching reactor in series with a microbial ferric iron regenerating reactor to mitigate microbial inhibition and to enhance base metal extraction. A mixed microbial culture dominated by L. ferriphilum, L. ferrooxidans and At. Caldus was maintained over a period of two months in the presence of polyurethane foam which acted as the biomass support particle. Microbial metabolic activity of immobilised cells and their tolerance to inhibitory Cu ions was investigated. Results showed that immobilised microbial cells maintained higher ferrous iron oxidation activity compared to planktonic cells in the presence of increasing Cu2+ concentrations up to 6 g/L. This presents an opportunity for microbial immobilised systems to be adopted in two-stage bioleaching system. Keywords: Bioleaching, microbial immobilisation systems, microbial inhibition and ferrous iron oxidation

Jan 1, 2020

By G. M. Evans, S. Mitra, A. V. Nguyen

As an alternative to the conventional flotation method, there has been continual research in an effort to enhance base metal ores and combustibles recovery using ultrasonics. These studies however were largely limited to the pre-treatment of ores for the removal of surface contaminants while the effect of ultrasonics on the bubble and pulp-phase interactions remained somewhat unexplored. Noting this deficiency, this study aimed to investigate the mechanistic effect of an ultrasonic field in an impeller less flotation cell specifically focusing on the bubble-particle interaction dynamics. A lab-scale flotation cell (capacity 1 litre) was designed, involving a vertically oriented ultrasonic probe (20 kHz, rated at 1.2 kW) and operated in the output power range from 1 to 20% with continuous liquid recirculation. Particle suspension was achieved by the combined action of recirculation and ultrasonic pulses. Interactions of coarse size glass ballotini particles (D32 = 314 μm) with bubbles (< 1 mm diameter) in the presence of a surfactant were studied, using high-speed imaging. Application of an ultrasonic pulse in the aqueous medium resulted in the generation of numerous microbubbles which increased with increasing input power and duration of the ultrasonic pulse time. Such periodic pulses introduced at a high level of turbulence in the pulp phase facilitated adequate bubble-particle contacts. Microbubbles played a key role in forming particle clusters which were observed to be floatable even in the absence of any carrier bubbles. When carrier bubbles were introduced, some of these particle clusters attached to the bubble surface and conveniently floated. It was noted that a favourable state of bubble-particle interaction was achievable to produce stable bubble-particles aggregates by suitably controlling the input power and period of the applied ultrasonic pulse. Keywords: Flotation, bubble-particle interactions, collision, attachment, detachment, ultrasonic pulse

Jan 1, 2020

By P. J. Bartlett, H. A. D. Kirsten

A two-dimensional numerical scheme for the simulation of the plastic stress-strain properties of rock, and the various components of rock reinforcement and shotcrete, was used to rigorously determine the characteristics of tunnel support under squeezing conditions. D-shaped tunnels varying in size from 2,8 to 6,5 m in height and width were considered. The four main mining stages in the BAS block cave on Premier Mine were simulated as alternative sets of field stresses. The rock reinforcement comprised 1,8 m long fully grouted rockbolts and 6,0 m long fully grouted cable anchors at various spacings, and of various tensile and bond strengths. Various thicknesses of shotcrete and footwall support configurations were also considered. It was found that shotcrete and rock reinforcement, owing to inherent differences in stiffness, do not function in unison. As monolithic application, shotcrete has an overriding avidity for the imposed loading and destructs itself unless of extraordinary thickness. Under squeezing conditions, it accordingly supports the excavation surface while the rock reinforcement contains on-going convergence. The loads in the various support components can be determined only by simulation of the shotcrete as a stress-relieved medium to represent its broken condition in the field. The loads in the rock reinforcement vary around a tunnel and are the largest in the sidewalls. The loads increase with an increase in tunnel dimensions, the increase in load in the sidewall due to an increase in height being larger than that in the hanging- and footwalls due to an increase in width. Depending on the strength of the rock, the loads in the rock reinforcement become prohibitively large beyond a certain size of tunnel. The reinforcement of footwall rock can be omitted without adversely affecting that in other walls, provided the tensile and bond strengths of the reinforcement are not limited. The resulting footwall heave may, however, result in undermining of the reinforcement in the adjoining walls. A relative depth of anchor of 1,37 provides adequate bond strength in the larger sizes of tunnel, whereas, in smaller tunnels, anchor length is determined by the allowable convergence. The lack of fun column grouting adversely affects the efficiency of rock reinforcement to a major extent. The loads in rock reinforcement are very sensitive to the strength of the rock. Owing to a lack in refinement and calculational rigour, empirical classification systems are an inappropriate means for the design of tunnel support in squeezing conditions and, as a result, grossly unclear-estimate the support requirements.

Jan 1, 1992

By F. K. Crundwell

The rate of dissolution of sulphide minerals is frequently interpreted in terms of the passivating behaviour of a layer of material on the surface. This interpretation has driven research to identify and possibly remove this layer from the surface, and thus yield higher rates of dissolution. The identification of such layers has proved difficult, with different researchers claiming different compositions for possible layers. An alternative framework has been pursued by the author and co-workers, who have argued that the rate of dissolution is governed by the electronic structure of the minerals which give rise to their semiconducting properties. Researchers sceptical of this approach have argued that semiconducting properties are only important in highly pure substances. In this paper, we review the electronic and semiconducting properties and discuss how they impact dissolution behaviour. These properties affect their colour and dissolution properties even though the minerals are not high purity substances. Keywords: Dissolution, leaching, passivation, semiconductor, chalcopyrite

Jan 1, 2020

By C. O’Donova, E. Adam, L. A. Torres-Cruz

High-profile failures of tailings dams illustrate the significant safety challenge they pose. The decant pond of such facilities plays an important role in their mechanical stability. This study investigates the feasibility of using public multispectral data collected by the Sentinel-2 satellite mission to monitor the area of the decant pond. The Sentinel-2 mission has a revisit time of 5 days and the data used herein has a spatial resolution of 10 m/pixel. We considered two water-detection metrics: the normalized difference water index (NDWI) and the near-infrared (NIR) reflectance. Implementation of the method at a South African platinum tailings dam illustrates the capabilities and limitations. Public, sub-metre resolution true-colour imagery served as validation data. There is good agreement between pond area measurements based on Sentinel-2 and the validation imagery. These results suggest the feasibility of using Sentinel-2 data to supplement the monitoring of the decant pond.

Apr 2, 2022

By K. H. Zhang, Y. P. Gao, J. Q. Zhao, Q. Li, J. C. Ma

Compared with sinusoidal oscillation, non-sinusoidal oscillation enhances the ability to control waveform and meets different oscillation requirements by adding an independent parameter, which is the key technology for high-efficiency continuous casting. A new waveform of mould non-sinusoidal oscillation is proposed in this paper, that is, a velocity waveform that is composed only of straight lines and parabolas, which not only has good dynamic characteristics, but also further enhances the feature of non-sinusoidal oscillation. Compared with sinusoidal oscillation, the negative strip time is reduced by 15.9%, the maximum negative strip velocity difference is increased by 38.3%, and the maximum positive strip velocity difference is decreased by 27.8%. The formulas of oscillation process parameters can be clearly deduced, and the isoplethic curves of oscillation process parameters when multi-parameter simultaneously changes are drawn. The negative strip time decreases as the modification ratio, frequency, and casting speed increase, but increases as stroke increases. In order to maintain the stability of negative strip time, an oscillation synchronous control model for optimising oscillation basic parameters was developed. Through practical application, the mean oscillation mark depths decreased by 10.1 μm and the flux consumption is increased from 0.37 kg/t to 0.45 kg/t in the non-sinusoidal oscillation model, which meets the requirements of normal production.

Oct 24, 2025

By M. Seerane, R. Machaka, P. Ndlangamandla

"Metal injection moulding (MIM) is a near-net-shaping powder metallurgy technique suitable for the cost-effective mass production of small and complex components. In this paper, the effects of the metal powder particle size on the final properties of 17-4 PH stainless steel are reported. Three different particle sizes (d50 33.0, 7.91 and 4.04 µm) were used to prepare the injectable MIM feedstocks using a CSIR-developed wax-based binder system at a fixed solids loading of 60 vol.%. The distribution slope parameter and the rheological flow index behaviour were used to predict the flowability of the feedstocks. The effects of the particle size on MIM products were established. The results are discussed in terms of the subsequent microstructural and mechanical properties of MIM 17-4 PH stainless steel against the established MPIF Standard 35. IntroductionMetal injection moulding (MIM) is a near-netshape manufacturing process established as an alternative to produce complex small-tomedium shaped components that were previously fabricated by conventional methods at higher costs (Attia and Alcock, 2011; Todd and Sidambe, 2013).Although recent advances in micromanufacturing (mainly machining and electrodeposition-based) have enabled the fabrication of complex micro-size geometries directly into miniaturized parts, these processes are generally neither routine nor easily accessible at present (Attia and Alcock, 2011).Typically, the MIM process consists of four steps: mixing, injection moulding, debinding and sintering. Mixing involves blending of the desired metal powder composition with carefully selected multicomponent organic binders at the right proportion. This mixture, which is called feedstock, is then injected into a mould of a required shape to form a ‘green part’. Debinding is the removal of the binders by chemical, thermal, or catalytic means while maintaining the shape of the part. Finally, the debound part is sintered to a full or near-full density part (Sotomayor, Varez and Levenfeld, 2010). The MIM process is essentially a combination of plastic injection moulding with conventional powder metallurgy technologies."

Jan 1, 2016

By H. Matsuoka, M. Kawata, K. Mitsuhashi, C. Tokoro, T. Kato

Copper sulfide minerals generally have high floatability, but their floatability decreases when surfaces are oxidised under some conditions. Sodium hydrosulfide, NaHS, is known both as an activator and as a depressant for copper minerals. NaHS at low concentrations activates copper oxides by sulfurising surfaces, but at high concentrations depresses copper sulfides by interfering with the adsorption of collectors. In this paper, we applied NaHS to activate oxidised chalcocite and bornite and focused on the conditions of NaHS treatments such as dosages and pH, and the effect of these conditions on the floatability of those minerals. The pure minerals treated with NaHS were analysed using zeta potential analyser and X-ray photoelectron spectroscopy. Analyses indicated that the reactions of NaHS and the products on the mineral surfaces after the reactions (i.e. oxidation, pH change) depended on whether NaHS reacted with copper or iron, and the reactions between NaHS and iron depend on NaHS concentration and pH. Flotation tests using a small column cell were performed with a single pure mineral, and using a Denver-type flotation machine with a copper sulfide ore. The results of each flotation test were utilised to clarify the floatability. The relation between flotation behaviours and analysis results were also discussed. Keywords: Chalcocite, bornite, NaHS activation, flotation, floatability

Jan 1, 2020

By M. Rudolph, A. van Hall, G. Öktem, S. G. Özkan, A. Hassanzadeh

Despite the four-decade study on ultrasound’s (US) impact on mineral’s floatabilities, there is still not clarity regarding its role in mineral surface wettability. For this purpose, the present investigation endeavours ……the hydrophobicity and floatability characteristics of a chalcopyrite-pyrite-quartz (Cp- Py-Qtz) system in the presence and absence of an ultrasonic pre-treatment. The ultrasonic process was carried out by a sonopuls at a constant frequency of 20 kHz with an adjustable power level from 30 to 200 W. Initially, the impact of sonication time (15, 30, 45, 60 and 90 s as well as 10, 20 and 30 min) and power magnitude (30, 60, 90, 120 and 180 W) were evaluated while the dissolved oxygen, temperature, conductivity and pH were also monitored. Collector-less micro-flotation tests were carried out on the non-pre-treated and the US pre-treated samples at 60 W and 15 s. The samples’ hydrophobicities were determined by the drop shape analysis approach. The dissolved-oxygen level was varied using a mini bench pressurised water reactor to study the effect of O2(aq) concentration on the chalcopyrite and pyrite wettability characteristics. The results showed that the minerals’ hydrophilicities were relatively more sensitive to the sonication’s time than its power that resulted in reducing all three minerals’ hydrophobicities. In addition, it was found that the dissolved-oxygen content and creation of sub-micron sized bubbles led to an improvement on chalcopyrite and pyrite hydrophobicities. Finally, we proved that the Cp’s floatability increased and Qtz’s recovery reduced while being subjected to the ultrasonic irradiation (15 s at 60 W); however, Py’s recovery remained constant. This conclusion confirmed the possibility of a selective separation in ultrasound-assisted copper ore flotation. We recommend that further advanced investigations be conducted, using an X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM) to thoroughly understand the surface modifications induced by the ultrasonication

Jan 1, 2020

By L. Linzer, J. Oelofse, T. Rangasamy

Extraction of isolated blocks of ground with elevated levels of stress and/or seismicity is a common occurrence in South African gold mines. Isolated blocks of ground that are divorced from the main orebody and are bound by complex, dip and strike parallel, non-persistent geological structures dominate the Klerksdorp gold fields. No quantifiable methods exist to assess the risk associated with mining these and other classes of remnants. A case history, from a gold mine that planned to extract three remnant blocks hosted in a large complex fault system, is used to illustrate the decision making process based on a combination of underground observations, numerical and seismic techniques. Simple risk estimation matrices were developed and used to estimate remnant-mining risk and finally to quantify the decision on which of the three remnant blocks could be extracted at the lowest possible geo-mechanical risk. Il est courant dans les mines aurifères sud-africaines d’extraire des blocs de terrain isolés qui possèdent des niveaux élevés de contraintes et/ou sismicité. Les régions aurifères de Klerksdorp sont dominées par des blocs de terrain isolés qui sont séparés du gisement principal et qui sont retenus par des structures géologiques non continues, à parallèle de pendage et direction, et complexes. Il n’existe pas de méthodes quantifiables pour évaluer le risque associé à l’extraction de ce genre et autres genres de résidus. Une observation de cas, celle d’une mine d’or qui avait l’intention d’extraire trois blocs de résidus qui se trouvaient dans un grand système complexe de faille, est utilisée pour illustrer le processus décisionnel basé sur une combinaison d’observations souterraines, et de techniques numériques et sismiques. Des matrices simples d’estimation de risque furent développées et utilisées pour estimer le risque lié à l’extraction des résidus et finalement pour quantifier la décision sur lequel des trois blocs de résidus pouvait être exploité au moindre risque géo-mécanique. Die Extraktion isolierter Gesteinsblöcke mit erhöhtem Spannugsniveau oder Seismizität ist in südafrikanischen Goldminen üblich. Isolierte Gesteinsblöcke, die von der Haupterzader getrennt und durch komplexe, parallele abfallende und horizontale nicht persistent geologische Strukturen gebunden sind, beherrschen das Klerksdorp Goldfeld. Es gibt keine quantifizierbare Methoden die das Risiko welches mit dem Abbau dieser und anderer Klassen verbunden ist beschreiben. Eine Fallstudie einer Goldmine, welche die Extraktion von drei in einem großem komplexen Verwerfungssystem eingelagerten Einschlußblöcken plante, wird verwendet um den Entscheidungsprozeß zu illustrieren, der auf einer Kombination von Untertag Beobachtungen und numerischen und seismischen Techniken basiert. Einfache Gefahreinschätzungsmatrizen wurden entwickelt und angewendet um das Abbaurestrisiko abzuschätzen und letztendlich die Entscheidung zu quantifizieren, welcher der drei Einschlußblöcke unter dem niedrigst möglichen geo-mechanischen Risiko extrahiert werden konnte .

Jan 1, 2003

By A. Muller

The magnesia-carbon refractories used in the ConArc process experience severe wear as a result of the varying slag composition during the blowing phase. This work determines the basicity range in which minimum as well as maximum refractory wear was experienced by the magnesia-carbon refractories during service. In order to determine an optimum slag composition, the crucible test in an induction furnace was used. The magnesia-carbon refractory bricks were cut into cylindrical crucibles to which a synthetic slag in the form of pellets was added. The basicity (B2) of the slag was varied from 0.5 to 3. Tests were done at 1 600°C for one hour. Apart from the crucible test in an induction furnace, visual inspection and SEM analysis were also done to investigate the corrosion mechanism. FACTSage® predictions about the interaction between refractories and slag were also done. These predictions were compared to the actual results attained. It was found that the extent of corrosion by slag attack decreases as the basicity increases. The most severe corrosion of the MgO-C refractories occurred at a basicity of 0.5 whereas the optimum basicity is above 2.0. SEM analysis indicated severe slag attack at the lower basicities with no visible refractory-slag interaction at higher basicities. TheFACTSage® predictions supported most of the actual findings. Keywords Magnesia-carbon, ConArc, blowing phase, slag resistance, induction furnace, crucibles.

Jan 1, 2011

By N. V. Nemchinova, A. A. Kozlov, A. V. Epiforov

The gold dissolution in sulfuric acid media during industrial oxidative leaching test of sulfide gold-copper flotation concentrates has been achieved. The reason for this was using technological water containing thiocyanates and cyanides, which can dissolve gold under these conditions. The carbon in column (CIC) process for gold recovery from these solutions was proposed. However, significant amounts of copper and iron were adsorbed onto the carbon during CIC technology industrial tests. This resulted in a decrease of the gold content in the loaded carbon up to 0.5 kg/t. Further development of the technology was conducted by optimisation of gold and impurities stripping and implementation of carbon reactivation. It was shown, that the copper extraction from the loaded carbon during cyanide treatment was about 95%. The copper content in the stripped carbon after cyanide treatment was 1.76 kg/t. The gold recovery during high-pressure elution by caustic solution was 98%. The hydrochloric acidic treatment with thermal pre-treatment (reactivation) for removing iron and calcium from gold-free carbon was proposed. In this case, the iron and calcium extractions from carbon were 40-45 and 50-55% respectively. These tests showed that for carbon recycling the sequence operations of ‘cyanide treatment, gold elution, thermal pre-treatment, hydrochloric acidic treatment’ can be used. Keywords: Thiocyanate leaching, recovery of gold, carbon adsorption, gold elution, copper stripping, thermal reactivation of carbon, activated carbon, carbon-in-column

Jan 1, 2020

By S. Liaghat, J. Rosenkranz, P. Semsari Parapari, S. Chehreh Chelgani, A. Tohry

It is commonly accepted that the energy consumption for comminution within mineral processing is responsible for around 3% of the global electrical energy usage. However, it was clearly noted that inside mills, less than 1% of this energy efficiency is actually consumed for particle size reduction. Therefore, monitoring of grinding parameters and modelling of product particle size distribution (e.g. characterised by the 80% passing size P80) play an essential role in operating costs of a mineral processing plant. Several empirical models have been introduced for the prediction of the P80. Since in the industrial scale usually a few numbers of grinding variables are monitored, these models cannot identically be used in plants. In addition, these empirical models do not always indicate the magnitude of correlations between milling variables. This study proposes to introduce the Hilbert-Schmidt independence criterion (HSIC) as an intelligent model for exploring linear and nonlinear relationships between grinding variables and P80 for an industrial scale data assessment. HSIC can rank variables based on their importance and help to select the most effective one for the generation of predictive models. HSIC modelling through various monitored grinding variables (ball charge, feed input rate, F80, power draw and work index) of an industrial plant indicated that F80 as a feed property has the highest effect on P80 of the ball mill products. Modelling results highlighted that the P80 can be accurately predicted, based on effective variables by using a HSIC model. These outcomes showed that HSIC as a powerful modelling tool can assess the relationships between monitored grinding parameters, select the most effective grinding parameters and satisfactorily predict the P80. Keywords: Product particle size, variable importance measurement, Hilbert-Schmidt independence criterion, mill charge

Jan 1, 2020