Search Documents

By M. Morales, J. Yianatos

"In this work, an alternative approach to evaluate the flotation performance along a flotation rougher row is presented. The collection zone as well as the froth zone performances were separately evaluated. For this purpose, sampling for metallurgical balances along with sampling of top-of-froth (TOF) grades in each cell of a flotation row were employed. The modelling of the rougher circuit consisted of using TOF grades for evaluating the collection zone recovery, because they are a good representation of the bubble loading, which identifies the true flotation process, as shown in previous studies for shallow froths. In addition, the feed characterization, size-by-liberation, allowed the identification of the mineral distribution and its evolution along the circuit. The latter result was obtained by fitting the recovery profiles as well as the TOF grades down the flotation bank. The flotation rates and the maximum recoveries in terms of size-by-liberation (size classes: -45 µm, +45-150 µm, +150 µm, and liberation classes: -20%, +20%-80%, +80%) were then estimated. The results were in good agreement with those reported in literature by other authors and showed the new approach allows flotation kinetics to be characterized in industrial circuits.INTRODUCTIONMineral recovery in industrial flotation circuits is a function of the mineral characteristics, such as particle size distribution and liberation (Trahar, 1981, Jameson, 2012), and operating conditions such as air flowrate, pulp flowrate, solids percent and reagent conditioning (Jameson 1977; Klimpel, 1980; Mehrotra and Kapur, 1974). The cell type and size (self-aerated or forced air) and the circuit arrangement (single cells or banks of cells in series) are also relevant factors. Sampling of circuits (e.g. rougher, cleaner or scavenger) are commonly carried out using standard automatic samplers for metallurgical balance assessment. However, the flotation plants are generally not provided with proper accesses for local streams, which hinders the recovery estimations and the flotation rate characterization."

Jan 1, 2018

By A. J. Lynch, E. V. Manlapig, J. S. Hall, G. C. Thorne

Modelling of industrial sulphide flotation based on data collected in several flotation plants is discussed. A mechanistic approach is used in which it is recognised that flotation is a two stage pro- cess with concentration occurring both in the pulp and in the froth. The important features of the model, which is a kinetic model, are – 1. for each particle species there is a range of flotation rates which corresponds to the range of particle sizes, 2. the flotation rates of the liberated gangue particles which have no hydrophobic properties are dependent on the flotation rate of water and are controlled by variables such as the air and frother addition rates, and the pulp level, 3. as an approximation the valuable sulphide particles may be regarded as being present in two groups with high and low flotation rate constants, 4. increase in the addition rate of collector tends to increase the proportion of sulphide particles which are in the high flotation rate constant group. The actual flotation rate of each group of particles is a combination of an "inherent" flotation rate and a flotation rate which is dependent on the flotation rate of water, 5. at particular air and frother addition rates and pulp levels there is a physical limit to the rate at which particles can be transported into the concentrate and this is recognised in the model. The model has been used to study the steady state and dynamic properties of flotation circuits in operating plants. The assessment of the behaviour of composite particles is difficult because of the problem of quantitatively describing the surfaces of mineral particles. A method which is being developed to measure unpolished particles is described briefly.

Jan 1, 1976

By J. Herbst, A. J. Lynch, E. V. Manlapig, G. C. Throne, J. S. Hall, G. Poling

Modelling of industrial sulphide flotation based on data collected in several flotation plants is discussed. A mechanistic approach is used in which it is recognised that flotation is a two stage process with concentration occurring both in the pulp and in the froth. The important features of the model, which is a kinetic model, are – 1. for each particle species there is a range of flotation rates which corresponds to the range of particle sizes, 2. the flotation rates of the liberated gangue particles which have no hydrophobic properties are dependent on the flotation rate of water and are controlled by variables such as the air and frother addition rates, and the pulp level, 3. as an approximation the valuable sulphide particles may be regarded as being present in two groups with high and low flotation rate constants, 4. increase in the addition rate of collector tends to increase the proportion of sulphide particles which are in the high flotation rate constant group. The actual flotation rate of each group of particles is a combination of an "inherent" flotation rate and a flotation rate which is dependent on the flotation rate of water, at particular air and frother addition rates and pulp levels there is a physical limit to the rate at which particles can be transported into the concentrate and this is recognised in the

Jan 1, 1976

By T Krull, T J. Donohue, D Ilic

"Problematic issues in materials handling applications are often amplified in the iron ore industry due to the relatively high density of the ore as well as the abrasive nature of the product. One of the main issues arising from the handling of iron ore is excessive wear, with two applications typically suffering accelerated wear issues being transfer chutes and conveyor belts. These wear issues are problematic due to additional capital costs, but mainly due to the associated downtimes resulting in the accelerated wear. This study will focus on the modelling of iron ore in practical applications where design work was undertaken in order to reduce the wear rates of different materials handling equipment.CITATION:Donohue, T J, Ilic, D and Krull, T, 2013. Modelling of iron ore in materials handling applications with a focus on wear, in Proceedings Iron Ore 2013, pp 355-362 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

Aug 12, 2013

By Y Chung, B Choi, H Lee

Numerous studies have explored refractory corrosion in the presence of slag and/or metal, utilising the Finger Rotating Test (FRT). Recent experimental results revealed that corrosion initially increased with rising rotation velocity but was then suppressed above a critical rotating speed, under conditions involving a two-liquid solution of slag and molten steel. This phenomenon indicated that the liquid/liquid interface movement occurred during spindle rotor. However, understanding of the phenomenon, known as the rod climbing effect, remains inadequate in the case involving slag, molten steel and refractories. To gain deeper insights, a simulation experiment was conducted on the phenomenon using silicon oil/water and the phenomenon was analysed using computerised simulation software (Ansys Fluent). Experimental modelling demonstrated that interfacial movement occurs even at low rotation velocity, as low as 200 rev/min, when the viscosity of silicone oil is at 100 cP. Computational analysis utilising the Volume of Fluid (VOF) method and a realisable k-ε turbulence model elucidated that interface elevation during rotor rotation arises due to viscosity disparities between the two fluids. Furthermore, this movement is more pronounced when the viscosity difference is lower. Additionally, the study findings suggest that higher density differences between two liquids suppress the phenomenon of interface movement.

Jun 19, 2024

By A. A. Hejja

"Liquidus temperature and electrical conductivity data measured on synthetic slags were modelled by the use of neural nets (NN). In this work the applied multilayer feedforward NNs were trained by a conjugate-gradient optimisation for which a three layer formulation was used. Very good fits were obtained for both the liquidus temperature and the conductivity data. The synthetic slags were prepared from pure oxides to represent a wide range of compositions likely to be encountered in ferromanganese and silicomanganese smelting. The slag constituents were in the following range: Mn0;5-30%, Ca0;20-35%, Mg0;5-15%, Si02;27-58%, Al20 3;5%. Liquidus temperatures varied from 1300° to 1380° and increased with increasing basicity ratio. The electrical resistivity of slags decreased with the increase of basicity ratio from 0.55 to 1.1 but above 1.1 basicity ratio the resistivity tended to increase depending upon the MnO content.Neural NetsVarious neural net (NN) topologies such as Hopfied nets, Hamming nets, Carpenter/Grossberg classifiers have been discussed in some detail by Lipmann (1). In this work, the applied multilayer feedforward NNs are not trained by backpropagation (BP) using the generalized delta rule (2) (GDR), but by a conjugate-gradient optimisation [3,4] - a procedure similar to that proposed by Barnard and Cole [5]. A neural net is capable ·of learning the non-linear functional relationship existing between a set of inputs and outputs [ 6]. In this capacity a neural net is used to determine the functional relationship that exists . between N dependent and independent variables, which characterize the system under consideration For a three layer neural net [1] the following function may be formulated, which relates X; inputs to the output Y 0 from node o:"

Jan 1, 1997

Modelling of Magnetic Carrier Separation Process Using Computational Fluid Dynamics

Sep 13, 2010

By Bo Zhou

During secondary aluminum recovery, complex aluminum scraps are melted and refined often in a rotary melting furnace and the aluminum metal has to pass through a molten salt layer and melt down in a bottom aluminum bath. Previous experimental study and industrial observations have indicated that salt shell formation and re-melting on the metal solids in the early stage plays a critical role in the overall melting process of the scrap. In the present paper, mathematical models are developed to simulate the process of salt shell formation and its re-melting, as well as aluminum melting in the molten melt. Influence of the salt layer properties, the residence time of the metal particles in the salt layer, the particle size, the temperature of bulk melt were investigated. Reasonable agreement with the measurement data was reached. The developed melting model will be used to construct a population balance model of the scrap melting behavior for a rotary melting furnace, taking into account the distributed characteristics of the scrap. Finally, an overall process model based on Computational Fluid Dynamics (CFD) for the rotary furnace will be constructed, coupling with the population balance model for the scrap melting.

Jan 1, 2003

By Bo Zhou

During secondary aluminum recovery, complex aluminum scraps are melted and refined often in a rotary melting furnace and the aluminum metal has to pass through a molten salt layer and melt down in a bottom aluminum bath. Previous experimental study and industrial observations have indicated that salt shell formation and re-melting on the metal solids in the early stage plays a critical role in the overall melting process of the scrap. In the present paper, mathematical models are developed to simulate the process of salt shell formation and its re melting, as well as aluminum melting in the molten melt. Influence of the salt layer properties, the residence time of the metal particles in the salt layer, the particle size, the temperature of bulk melt were investigated. Reasonable agreement with the measurement data was reached. The developed melting model will be used to construct a population balance model of the scrap melting behavior for a rotary melting furnace, taking into account the distributed characteristics of the scrap. Finally, an overall process model based on Computational Fluid Dynamics (CFD) for the rotary furnace will be constructed, coupling with the population balance model for the scrap melting.

Jan 1, 2003

By E Werner, J J. Donovan, A J. Morris

Coal mining in the Pittsburgh coal seam, in the eastern United States, began in the late-1700s. Since that time numerous mining technologies have been used throughout the basin. Environmental concerns, particularly mine discharge quality, have only been considered since the early-1970s. During the last 20 years, many of the larger underground mines have closed and are now either flooding or already flooded. In order to understand the consequences and potential future effects of this flooding, mine mapping of the basin has been compiled utilising a geographic information system (GIS) with an emphasis on locating barrier pillars between mines. Mine flooding has been monitored at numerous sites utilising pressure transducers and data loggers. These data have been combined with historical data from mining companies and structure contour maps to yield chronological flooding maps of the basin from 1980 to present. Utilising these maps and the indicated flooding rates, future discharge locations have been identified along with estimates of when these discharges are likely to occur. Related work has identified barrier pillar hydraulic conductivities and vertical infiltration rates as a function of overburden thickness. All of this information has been integrated into a groundwater flow model that simulates mine flooding. This model is useful in defining mine to mine interactions as the various mines fill with water, and the testing of various flooding scenarios. The model will be used in evaluating mine discharge rates and evaluating the options for regional mine drainage treatment.

Jan 1, 2003

By A Kamkar-Rouhani

The paper describes a mineral potential mapping of gold and other base metals in Mahneshan area, Iran, using geographical information system (GIS), in which, the method of weights of evidence is used to combine spatial data obtained from different sources to make prediction for mineral potential. This is a quantitative tool for mineral resource mapping that can be used to delineate and predict favorable areas for further exploration by using the characteristics of known and similar mineral deposits. In accord with the descriptive model for gold and other base metal deposits, the spatial distribution of 20 known mineral occurrences was used. In addition, digital evidential maps were used in this study. The maps include maps of lithological units, faults and first derivative of total aeromagnetic data, remotely-sensing data of alteration and also geochemical evidential maps produced on the basis of the results of chemical analyses of samples taken from the stream sediments in the area, and analysed for Cu, Pb, Zn, Sb, As, Ag, Au, Ba and Hg elements. The binary maps were produced and then combined with the model of weights of evidence using 20 known mineral occurrences to obtain a final map of the favourable areas for further exploration of gold and other base metals in the study area. The final map is a predictive GIS model of the mineral deposits in the area. An interesting matter about the final predictive GIS model is that the faults are mainly located in favorable areas which possesses the most probability of mineralisation. Therefore, the faults in these areas have priority for further exploration of gold and others base metals.

Jan 1, 2005

By P B. Hills

This paper discusses the computerisation of the drill hole database, some historical records and the creation of a three-dimensional model of the gold bearing structures for resource estimation and mine design at the Beaconsfield Gold Mine in northern Tasmania. The phase of the work reported here commenced in November 1997 and was completed in March 1998. The historical mining on the Tasmania Reef had reached a depth of 450 m below surface when work ceased in 1914. Production up to this time was 854 000 ounces gold. Recent activity has concentrated on testing depth extensions of the previously mined ore structure, to a depth of 1000 m below surface, and the establishment of an underground mining operation to exploit this resource. The lode structure at Beaconsfield is a typical mesothermal vein style deposit. Gold is contained within narrow quartz, carbonate and minor sulphide reefs or veins. Reef width is variable from a few centimetres to several metres with æsplitsÆ or bifurcations being common. The reef averages 2.7 m true width in the economic zones. The key feature of this style of mineralisation is that reef continuity can be confirmed by drilling to a high level of confidence, but the gold grades are somewhat more erratic in nature. The resource estimates were based entirely on drill data. The drill intercepts on the reef were coded for each split directly into a database and these allocations validated graphically. The distribution of grades and widths show an increasing gold grade with reef width for both the drill hole and historical data. The same histogram patterns, mixed populations and overall arithmetic average grade for the historical mine (at 21.52 g/t > 3 g/t Au) and new drill data (at 21.55 g/t > 3 g/t Au) would indicate that there has been no major change in the overall style or grade of the ore from the historically mined areas to the new resource. The reef, host geology, area of historical workings and resource zones at the Beaconsfield Mine were defined as a series of closed wireframes. The Tasmania Reef was modelled as five mutually exclusive wireframes, one each for the Main Reef, Footwall split, Second footwall split, Third footwall split and Hangingwall split. The data for this exercise was extracted directly from the drill hole database, stored in Microsoft Access format. The formation of the solids was largely automated with manual validation and refinements allowing for multiple runs using different interpolation methods and combinations of drill holes. A rotated 3D-block model in the plane of the Tasmania Reef was created to contain the grade model and allow for tonnage, grade and reef width estimates. An inverse distance squared (ID2) interpolation method was used with ranges of 25 m along strike and 125 m down plunge with blank areas infilled with 50 m by 250 m. Each reef split was filled separately using only drill intercepts from that split. The method was independently validated used MIK methods. Longitudinal sections have been prepared showing gold grade, reef width and width +grade. Estimates were also made for arsenic, sulphur and other minor elements. The long sections show three distinct mining zones. A method is presented for the conversion of the resources to reserves by the addition of mining parameters for minimum mining width, minimum pillar, cut-off grade and internal and external dilution. The selection is done on columns within the block model, and allows for all splits to be considered at once. Multiple runs can be made for varying combinations of parameters to test different mining methods and their impact on reserves, mine life and NPV.

Jan 1, 1999

By S Bhandari

This paper is a step in the direction of developing a computer model of the near source dispersal of suspended particulate matter (SPM) in planetary boundary layer (PBL). Since Surface Mine blasting is done during daytime, study of daytime dispersion in convective boundary layer (CBL) is relevant to the mining engineers. The scope of this paper covers dispersal after the blast in any type of PBL ie stable, convective and neutral but only under weak wind condition. Mathematical analysis of the dispersion of vertical dust column has been carried out during weak wind conditions in a tropical meteorological environment. It is noted that dust concentration reaches its maximum value some distance away from the point of blast, immediately after the blast (t*=1) and thereafter rapidly decreases. After a short interval from the time of blast at t*=11 the dust concentration shows linear decrease in concentration with distance. Dust concentration also rapidly drops during the interval t*=1 to t*=11, thereafter, rate of drop is nearly the same at any distance from the point of blast.

Jan 1, 2001

By M. Kozlowski, I. Candy, S. Kashani-Nejad

The production of ferronickel and matte from nickel laterite ores is dominated by the Rotary Kiln, Electric Furnace (RKEF) process. Over time, major improvements have been made to kiln operation. In particular, kilns are now required to perform substantial amounts of reduction of the metallic oxides, thereby reducing the load on the electric furnace and substituting less expensive fossil fuel for electrical energy. To the basic refractory lined cylinder have been added: internal features such as dams and lifters, aimed at changing residence times in different sections of the kiln and enhancing gas/solids interaction, on-board fans to introduce tertiary air and on-board scoop mechanisms to allow reductant coal additions to be made at advantageous locations along the kiln. Many of these advances in design have been made on a trial and error basis and by 'borrowing' technology from related kiln processes, supplemented by proprietary modeling to various levels of sophistication. This paper presents a conceptual review of modeling components for the development of a computational model for granular material flow, combustion and heat and mass transfer in a nickel laterite kiln. These components include: granular material flow and segregation models for the simulation of bed motion, heat transfer models that simulate heat transfer by turbulent diffusion and radiation and the distribution of this thermal energy within the bed, and fluid flow models for the prediction of combustion aerodynamics.

Jan 1, 2005

By M. W. Bates, A. L. Mular

This paper, based on work carried out at Queen's University, describes the development and use of mathematical models of the parallel cyclone classifiers being used in the Strathcona mill of Falconbridge Nickel Min.es Ltd.

Jan 1, 1971

By G. P. Demopoulos

The solubilities of calcium sulfate in HCl-CaC1z-H20 solutions up to 100°C were successfully modelled with the aid of OLI software after bringing proper modification to it. Modelling, more specifically, was carried out with the help of the Environmental Simulation Program, V-6.6.0.4 (www.olisystems.com). New model parameters were obtained by regressing the solubility data. A private databank was established by adding the newly defined model parameters and the calcium sulfate hemihydrate phase that is not included in the original OLI database. The McGill-modified OLI software was successfully applied not only to modelling solubilities but also the transformation of the various CaS04 phases, namely dihydrate, hemihydrate and anhydrite. In this paper, the modelling methodology will be discussed along with a comparison of the capabilities of the newly developed model vs the original OLI model.

Jan 1, 2005

By Douglas Morrison

By applying modern 3D numerical modelling techniques to the analysis of a pillar cascade failure in a room and pillar mine, it has been possible to simulate the collapse and develop an understanding of the role of low rock strength areas within the mine. The cascade collapse occurred in a normal strength area but would not have occurred without the influence of a low strength rockmass in one section of the mine. The numerical analysis made it possible to develop a realistic assessment of the residual pillar strength. With the introduction of regular empirical pillar performance assessments, critical pillars or groups of pillars can now be identified proactively. This has led to an aggressive extraction ratio of 90% that manages pillar failure, avoids pillar and roof collapse and prevents surface subsidence. Since the mine also has significant waste products, the use of backfill for ground support has made it possible to significantly improve the performance of strategically located pillars, and so achieve even higher extraction ratios.

Jan 1, 2003

By A D. Campbell, G R. Power

The increasing availability of marker data in sublevel caving operations is changing the way draw can be managed to achieve positive recovery outcomes in these mines. This paper describes how marker data has been incorporated into flow simulations to help calibrate full mine ore recovery simulations for the forecasting and optimisation of recovery.Some flow modelling techniques do not use marker data, relying instead on the outcomes of physical and numerical experiments to guide assumptions about an idealised recovery flow shape. Other mines have access to marker trial data, but find it difficult to specifically link this marker trial information to their recovery models. This can make it difficult to create a link between the marker trial information and recovery forecasting and limits the potential value of the marker trials at the mine.In the simulations shown here, the marker rings are fully incorporated into the flow model, and can be tracked in the model. The model is compared against the real underground marker recovery information to give the user an indication of how accurately the flow settings in the model are reflecting the conditions in the experimental marker rings. An example of use of this system at an operational mine is included.CITATION:Power, G R and Campbell, A D, 2016. Modelling of real-time marker data to improve operational recovery in sublevel caving mines, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 105–110 (The Australasian Institute of Mining and Metallurgy: Melbourne).

May 9, 2016

By J. Yianatos

The regrinding stage is necessary to achieve the particle size (liberation) for the final upgrade in the cleaning flotation process. At present, vertical mills (Vertimill) are employed at industrial scale, using small steel balls as grinding media. The regrinding process characterization requires mixing regime knowledge along with the mean residence time of the liquid and solid phases. The residence time distribution (RTD) and the actual mean residence time estimations from theoretical relationships are complex due to the presence of three components (solid mineral, balls and water), mass transport characteristics and segregation affecting solid holdup. Therefore, the experimental RTD determination is suitable for industrial mills. In this work, RTD measurements were performed in two different industrial flotation plants. Both plants operate with Vertimills of 1000-1500 HP in the regrinding process with different circuit arrangements according to the plant treatment capacity. The RTD was measured using the radioactive tracer technique. This technique allows for non-invasive tracer detection where the procedure consists of selecting a liquid (Br82 in aqueous solution) or solid (irradiated solid) tracer for RTD data acquisition in real time. The RTD was modelled by the Large and Small Tanks in Series (LSTS) model, which allowed the characterization of the different mixing regimes for both liquid and solids in the regrinding mills. Effective mean residence times in the range 2-11 min were obtained with RTD shapes similar to those obtained in tumbling ball mills. It was found that residence times of liquid and fine particles were similar, which is consistent with results observed in other mixing systems, e.g. mechanical flotation cells. Also, the mean residence times estimated from the RTD allowed the mass balance adjustment to be validated from metallurgical sampling data. RTD estimations are useful for better understanding the Vertimill behaviour and to identify the actual breakage rates. Keywords: vertical mill, residence time distribution, particle size, modeling

Sep 1, 2012

It is desirable to have a model, or method, for major project development, which will enable flexible modelling of separation circuits. This ability improves the understanding of circuits and how they should be operated, especially in a design phase, where various options are evaluated and decisions are made on the required number of stages and configuration. This paper describes existing gravity separation circuit modelling theory and concepts, as defined by Holland-Batt1 (separation window, separation efficiency and proportionality of efficiency). The crux of the paperis the application of numerical analysis to fit raw separation data by means of a modified Bézier-Spline. The paper proposes an alternative to the theory developed by Holland-Batt that provides more flexibility in the modelling of separation stages and curves, which include S-shaped curves. The paper proposes a solution to the challenge of continuously extrapolating data, and the first difficulty normally encountered, which is the fitting of an appropriate empirical formula to raw data.

Jan 1, 2007