Search Documents

By Carmina Quintanar, Mario Palominos

"The mining industry is currently facing significant challenges to identify and adopt sustainable collector chemistry that effectively recover value minerals in difficult-to-process low-grade resources. These are typically characterized by complex mineralogy and the presence of significant amounts of penalty gangue sulfide minerals, especially pyrit which is a challenge to separate from value minerals at many major mine operations. Three chemical strategies for dealing with gangue sulfides and penalty elements are: a) selective flotation of value minerals while rejecting penalty minerals in the whole circuit, b) rejecting penalty minerals in an appropriate part of the circuit using selective depressants, c) using a combination of selective collectors and depressants in appropriate parts of the circuit. This paper will evaluate different strategies for dealing with those penalty elements and propose the most efficient alternative. New products and application technologies have been developed in the past 25 years by Cytec for implementing these strategies as dictated by the particular needs of a given plant. However, in recognition of the growing interest in meeting the sustainability challenges, Cytec Research and Development have been focused in the development of greener products (collectors, modifiers and frothers) and processes using FLOTATION MATRIX 100TM approach used to develop the optimal collector system for each specific ore. An overview of these new developments is given in this paper using case studies of application of selective collectors and polymeric modifiers including the newer, greener chemistries."

Jan 1, 2014

By Ji Hoe Kwon

Hindered-settling column process has been used for the separation of mineral particles for many years. In this study, a model based on DEM (Discrete Element Method) was used to simulate the movements of particles in a settling column separation process. DEM model is capable of describing the motion of particles and calculating the impact forces between particle and particle, and particle and wall contact under various operating conditions. Drag force, rotation induced Magnus lift force, contact force and the gravitational force acting on individual particles were considered for establishing the mathematic models. The overall motion of the particles in water is satisfactorily simulated by the DEM, and the result shows good agreements with the result of lap-scale experiment.

Jan 1, 2014

By R. P. Bhagat

The paper reports results on sinterability of iron ores from the mines of Tata Steel Limited. The objective was to improve tumbler index (TI) of sinter and its reduction parameters, reduction degradation index (RDI) and reducibility (RI) while maintaining strand productivity same or to improve. Pot sintering experiments were carried out with Joda and Noamundi mines having LOI 3.8% and 2.9% respectively. Sinter basicity (CaO/SiO2) was varied in the 2.1 to 2.5 range and its MgO content was kept in the 1.4% to 1.8% range according to the blast furnace requirement. The ignition condition and vacuum was maintained according to the plant practice. The effect of variables namely, granulometric composition of ores vis-à-vis that of sinter mix, bed height, solid fuel, moisture content of green mix on the sintering indices was studied. The productivity with Noamundi ore (26 %+6.3 mm) was lower than that with Joda ore (45% +6.3 mm) at 4% coke in the mix and sinter basicity (CaO/SiO2) 2.1. However, with 4% coke (ash 22%) the RDI value was significantly high in case of Noamundi Ore, which decreased significantly to 29% with 5% coke. The size parameter of ore fines and its LOI content do influence the sintering speed and the sinter properties significantly. At 2.1 basicity ratio, the sinter with the Joda ore having coarsergranulometry had highest TI (=70.2%). The RDI (27%) in this case was inferior to that (23.2%) when Joda ores with finer size was sintered. Better RDI with finer sized ore could be attributed to better assimilation of CaO/ MgO with more balanced and uniformly distributed hematite minerals. The size parameter has been inter-influenced by the coke requirement. With the lower mean sized Joda ore the sinter RDI decreased significantly to 23% and the productivity improved to 42.9 T/(sq m*d) when the coke in the mix increased from 4% to 5%. An attempt to reduce the coke in the mix from 5% to 4% with respect to both the ores was not successful as it resulted higher RDI and lower strand productivity. The present study has revealed that sinter RDI improves (decreases) with decrease in the sintering time, while such correlation was absent between TI of sinter and speed of sintering. Corollary to this observation, one can take advantage of other stronger parameters to increase the sintering speed (and hence sinter productivity) without affecting the sinter strength. In the present study higher FeO sinters has displayed excellent RDI, however cold strength of sinter was not correlated with its FeO content. The TI and RI parameters of sinter improves marginally with increase in its MgO content while RDI got worsened. Keywords: LOI content, iron ore, reduction properties, sinter, noamundi, joda

Sep 1, 2012

By Jacek Kolacz

A new sensor based sorting system has been developed, based on complex image analysis, where an image texture and a shape pattern is considered for identifying particles for separation. The sorting system employs different images generated by visual light and X-ray to provide maximal information about processed particles. Application of X-ray makes it possible to analyse the internal structure of the particles, which provides a 2D image with complicated texture, however, which contains valuable information. The advanced system for the texture and the pattern recognition is applied to analyse these images simultaneously, to provide the required identification of the particles for efficient separation. The mathematical models applied during identification and separation, are assisted by high speed calculations being specially optimised for this kind of process. This allows sorting with high capacity and high accuracy, which is critical for practical applications in mineral processing. The paper describes a number of tests during sorting of iron ore and coal with high separation efficiency.

Jan 1, 2014

By Sanjay Agarwal

Polymetallic sea nodule is a good source of manganese having 22.3% Mn, 5.4% Fe, 6.14% Si, 1.1% Cu, 1.15% Ni, 0.076% Co & others. These nodules are available in plenty on the sea beds, generated over thousands of years. Nodules mainly consist of manganese dioxide and iron oxide phases and other metals are entrapped in the complex cage of these Mn-Fe oxides phases.World-wide research is going on sea nodules as an alternative future source of many metals, which has led to intensified efforts to develop practical and economical processes for recovery of these metals. Very little attempt has been made world over on manganese extraction from sea nodules. We at CSIR-NML, India have developed a direct reduction smelting process of nodules to recover manganese as Fe-Si-Mn alloy. This alloy is extensively used in the steel industry as deoxidizer and as an alloying element to make different grades of steel. Lab scale experiments were conducted to study the manganese distribution between Fe-Si-Mn alloy and FeO-SiO2-MnO-CaO-Al2O3-MgO slag at different temperatures by varying the basicity of the slag. The direct reduction smelting of nodules is carried out on 1 kg scale in 50 kVA electric arc furnaces to have Cu, Ni and Co alloy and Mn rich slag. This manganese rich slag is further treated in 50 kVA arc furnace to produce Fe-Si-Mn alloy. The composition of Fe-Si-Mn alloy obtained is in the range of Mn: 60-65%, Si: 14-17%, C: 1- 2%, S: 0.03% and the balance iron which is acceptable grade for industrial use. The equilibrium amount of MnO in the final slag is controlled by the temperature of the final melt and the silica, lime, alumina content of the final slag. Increasing the basicity of the slag plays a major role in minimizing the loss of MnO to slag, increasing the amount of SiO2 of the slag, increases the manganese loss in the slag. Increasing the CaO/ Al2O3 ratio of the slag decreases the manganese loss in the slag. Increasing the MgO/CaO ratio of the slag increases the manganese loss in the slag. Thus by adjusting

Sep 1, 2012

By Sanjeeva Latchireddi

The effective utilization of the energy in comminution devices depend on how efficiently the broken particles leave the breakage field as soon as they reach product size. In crushers the transport of broken particles smaller than CSS is done by gravity (vertical) thus allowing the energy to be applied to new particles. However, in AG/SAG mills, gravity becomes the energy field and the transport of particles is horizontal which needs a suitable discharge mechanism and has been found to have significant effect on grinding efficiency. The material transport in AG/SAG mills depend on how efficiently the discharge system (grate and pulp lifters) operates. Material transport plays a significant role on efficiency of horizontal grinding mills ? AG/SAG/Ball mills. The magnitude of benefits that can be achieved with efficient material transport in AG/SAG mills has been demonstrated in a Greenfield milling operation at Mirabela Mineração?s SantaRita Nickel operation located in northeastern part of Bahia in Brazil. This grinding circuit was commissioned in October 2009, which was designed as a typical SABC circuit to grind 4.6 mtpa (575tph) to produce a P80 of 125um with 10-12% ball charge in SAG mill. However, it has been operating as ABC circuit without any grinding media in SAG mill. The nameplate design capacity has been achieved in few days and has gradually increased to an average of +800 tph and still it is operating in ABC mode. The milling circuit commissioning resulted in small and relatively simple measures to rectify, which was reflected in the rapid commissioning of the milling circuit and it achieved the nameplate capacity within days. The Santa Rita grinding circuit consists of one 30?x 15.4? 8.5MW SAG Mill fitted with Outotec patented Turbo Pulp Lifters (TPL TM ) in close circuit with HP400 short head pebble crusher and one 20? x 28.4? 5.5MW Ball Mill in closed circuit with Warman cyclone cluster fitted with 10 x 26? CAVEX cyclones. This paper examines the comminution design process, operating philosophies, experiences, post commissioning optimization, and future development works. Evaluation of the milling circuit performance relative to the design criteria is also discussed. Keywords: Autogenous Grate (AG), Semi-Autogenous Grate (SAG, Turbo Pulp Lifter (TPL)

Sep 1, 2012

By Zhazira Baltabekova, Nina Lokhova, Arailym Dukembayeva, Madali Naimanbayev, Zhantore Dzhurkanov

"Recycling of industrial waste provided they have enough valuable elements requires significantly less investment compared with ore processing. Promising man-made raw material of rare earth elements (REE) is phosphogypsum. Works performed on dismantlement of phosphogypsum only concern phosphogypsum processing from apatit. In this case, none of the developments was tested for phosphogypsum produced on the basis of phosphorus, chemical and mineralogical composition of which is markedly different from apatites. We studied phosphorites of the Karatau deposits (Kazakhstan). Mineralogically all the layers are composed of three main ingredients: phosphate, silica and carbonate, which account for 92-96% of the weight composition of ores and rocks. Of impurities there are pyrite, iron hydroxides, gypsum, fluorite. In addition to the actual composition of phosphate in phosphate ores there are apatite. Silica is presented by chalcedony and quartz, and carbonate by calcite and dolomite. The results of X-ray microanalysis showed point cluster of grains containing REE in samples of phosphate ore. Analysis of the grain showed that REE are in the form of phosphates and silicates, which according to numerous studies in apatites is not established. In this regard, phosphogypsum, waste of production, is different in composition too. Therefore, transfer of techniques proposed in the methods for extracting REE from phosphogypsum after processing apatite, did not give the stated results and required a new approach to solving this problem. It should be noted that the phosphogypsum produced by processing the Karatau phosphorite contains medium group of REE by 2.6, and heavy group by 7.8 times more than phosphogypsum from processing the Kola apatite. Given the shortage of medium and heavy REE research to develop ways to extract the lanthanides from phosphogypsum after processing phosphate rock is promising. The study contains results of studies on the extraction of REE from phosphogypsum after Karatau phosphorite processing."

Jan 1, 2014

By D. DePaoli, P. Zhang

"Phosphate ore is one of the most significant secondary resources for rare earth elements (REEs), and Florida phosphate is of particular interest due to its much higher ratio of heavy REEs, especially yttrium. Analyses of numerous samples from different phosphate processing streams in Florida showed that heavy REEs account for about 50% of the total REEs in Florida phosphate with yttrium accounting for approximately 30%. It was estimated that the amount of yttrium from US phosphate processing could meet the world demand for this critical element. This research aimed at recovering both REEs and phosphorus from a phosphate flotation tails. This tails is composed of mainly silica and phosphate with a small fraction of heavy minerals including monazite and xenotime. Since the specific gravity of silica is 2.6 versus 3.2 for phosphate and 4.5-5 for rare earth minerals, pre-concentration of the minerals of interest by gravity separation was investigated first. A pilot scale shaking table was used in this study. One pass gravity concentration achieved a shaking table concentrate analyzing over 600 ppm REEs at about 40% recovery. Flotation of the shaking table concentrate resulted in a product containing 1809 ppm REEs and 26.37% P2O5 with corresponding recoveries of 73% and 82%. Sulfuric acid leaching at 75oC could recover over 60% REE and nearly all the P2O5. Higher REEs (>90%) recovery was achieved by conducting leaching at higher temperature with concentrated sulfuric acid. The phosphoric acid from leaching of the flotation concentrate is of superior quality compared with that generated from regular phosphate rock.INTRODUCTION Rare earths are a group of 17 elements including those of the lanthanum series, as well as the non-lanthanides, yttrium and scandium. These elements have very specific critical uses in a multitude of markets. For example, lanthanum is used as a catalyst in breaking down crude oil to produce gasoline, diesel and jet fuel and neodymium is the core magnetic component in high-strength magnets for electric car batteries, wind turbines and hard disk drives. Rare earths are also essential for many military applications. Many of these applications have no substitute materials and the move to green technologies has dramatically increased their demand."

Jan 1, 2018

By K. S. McClure, S. B. Chitralekha, R. B. Gopaluni, Chmylek. T., S. L. Shah

This is a tutorial paper on industrially successful approaches for nonlinear process monitoring. Online process monitoring is essential to continuous operation of plants at high efficiency. The problem of process monitoring is one where an impending but undesirable process condition has to be identified and used to alert process engineers. This problem has traditionally been solved by practicing engineers using statistical techniques that detect a change in the underlying correlation structure of process variables. For linear correlation structures, algorithms such as principal components analysis, independent component analysis and partial least squares have been proposed. However, many industrial processes exhibit nonlinear correlation structure between the process variables. In this tutorial we describe various nonlinear monitoring techniques and elaborate on two recently developed methods, namely Support Vector Machine and Local Linear Embedding. The two methods are evaluated by application on two experimental case studies, a SAG mill and a polymer extruder.

Jan 1, 2014

By G. M. Wolfaardt, B. C. Qi

"From a lifecycle point of view, biologically digested sludge is a valuable product that should be reused if possible. Sludge may be applied to agricultural land for further sludge treatment and improvement of the characteristics of the soil. The characteristics of sludge that affect its applicability for soil conditioning include moisture content, pathogens, nutrients and metals. In this paper, the use of solid anaerobic digestion sludge of the lignocellulosic residues to condition soil was investigated in regards to nutrient composition, water content/retention capacity, hygienic quality (contamination of pathogenic bacteria) and heavy metal adsorbing/releasing characteristics of the sludge.It was found that the digested sludge had approximately 44.5% free and immobilized water and 46.9% bound water. The water-holding capacity of the sludge was high. This was a merit to apply the sludge to agriculture in dry areas. Anaerobic digestion under thermophilic conditions showed a promising disinfecting effect to the sludge, in contrast to the landfill digestion. The anaerobic digestion sludge could be considered as a humus-rich hygienic product that improves soil fertility and water holding capacity, nourishes plants and immobilizes heavy metals in the environment"

Jan 1, 2003

By G. Jenny

The particle size distribution of ground minerals is of decisive importance for downstream processes like floatation, separation, drying or others. In order to optimize the process efficiency on one hand and the final product quality on the other hand an early control of the grinding or classification stage is essential. The ultrasound based particle size analyser OPUS is capable of performing on-line particle size distribution (PSD) and solid content measurement of mineral slurries such as iron ore, copper ore, fluorspar, or quartz sand and alumina directly in heavy and rough process environment. The instrument provides a PSD of 31 size classes and can be applied for particles from nanometer range up to several millimetres. Arbitrary interpolations in between the 31 measured points are available. Typical residue values show an excellent agreement to off-line particle size analysis methods such as e.g. sieve analyses or laser diffraction. OPUS is designed as rugged in-line probe made from stainless steel. The applied technique is independent of product flow dynamics, contamination, optical properties, vibrations, pressures or changes in solid concentrations. The instrument provides all data on the system?s PC or on the process control system by a variety of interfaces.

Jan 1, 2014

By David G. Hulbert, Daniël J. Oosthuizen, Bernard Muller, Daniël de V. van der Spuy

"Expert Online is a software system that makes use of novel artificial intelligence technology, known as the Interpreting Expert System (IES), to mimic an informed expert on a minerals processing plant. Previous implementations of the IES were only capable of explaining the controller actions. Expert Online will also take on the more difficult task of explaining the complex process behaviour to the operator in simple language.The primary functions of Expert Online are to serve as a training tool and an online assistant to operators. However, it is also expected to be of value to supervisors and plant managers.A number of additional features are envisaged, including different operating modes, fault detection, speech synthesis and advanced message filtering. Expert Online will increase the operator’s knowledge of the process and his/her understanding and acceptance of the complex control systems. Users would be supplied with relevant and timeous information required to make informed decisions to ensure smooth plant operation.This paper discusses the structural design of Expert Online for implementation on the latest version of MINTEK’s open control platform, PlantStar, as well as several advances and additions since the previous implementations of the Interpreting Expert System."

Jan 1, 2003

By R. Venugopal

The paper reviews the development of Mineral Engineering education and training programmes in India. It highlights the major academic programmes and describes the evolution of the curriculum and design of courses in response to the changing requirements of India?s mineral and coal industries. An outline of the efforts needed to sustain and further improve the existing programmes is also provided. Keywords: manpower, mineral engineering

Sep 1, 2012

By R. K. K. Mbaya

The concept of adsorption or separation of platinum anions from base metal cations in chloride media using native or chemically modified activated carbons has been discussed in various journals within the context of batchwise systems. In our research work, a fixed-bed column, loaded with native and amine-immobilized activated carbon pellets (Norit RO 0.8), was used to study selective adsorption of platinum ions from chloride media at different flow rates, bed depth and initial metal concentrations. The kinetics, mass transfer characteristics, breakthrough data and separation factors (ß(Pt/Fe) and ß(Pt/Cr)) were quantitatively analysed and modeled. The bed-depth service time (BDST) model was used to model the breakthrough curves successfully in fixed-bed columns and also the dispersive diffusion in the radial and axial directions of the bed were evaluated. The inner diameter of the column was 3 cm and the bed depth was between 1.0 cm and 2.5 cm. Concentrated 4 M HCl acid solution was used to elute the fixed-bed column. It was demonstrated that amine-treated activated carbons provides superior separation factors between platinum and base metals (Fe, Cr). Results showed also, that 95% saturation was reached within 10~26 hours depending on the nature of ACs at a flow rate of 11 mL/min and initial concentration of 5 ppm.

Jan 1, 2014

By B. Csoke

As opposed to large household appliances, in the processing of which dissambling has an important role, this possibility is pushed into the background as the size decreases. The practicable solution is physical liberation of the used small electronic household devices (by comminution) and separation of the liberated components according to their physical properties. In the first phase of the experimental work the structural/material composition of the typical products (mixer, kitchen robot, hair dryer, coffee machine, iron, bread toaster, vacuum cleaner, oil fryer) was determined. In the second phase a physical liberability test for each appliance using a rotary shear/shredder for crushing and a hammer shredder for milling was carried out. In the third phase the experimental separability tests using an air classifier and a shaking air table, as well as an eddy-current and an electrostatic separator were performed. In the present paper the results obtained using the air classifier and the electrostatic separator, are reported. To follow the life of a grain a new methodological procedure was elaborated. The particle size fractions after comminution were sorted into material fractions (copper, aluminium, plastics, etc.), then separated, and finally the recovery of the material components in the products were determined. Furthermore, the joint separability was investigated from two points of view: 1) for grain fractions and 2) for each waste appliance. According to the tests carried out so far, it was established that it is reasonable to form appliance groups by selection (e.g. mixer, iron, bread toaster and similar appliances form the first group, while the sandwich oven,

Jan 1, 2014

By B. Das

The processing of banded iron ore by direct anionic and reverse cationic flotation to produce pellet grade concentrate is investigated. This low-grade iron ore, known as banded hematite quartzite (BHQ) contains 41.8% Fe, 41.5% SiO2, 1.2% Al2O3, and 0.5% LOI. The flotation behavior of BHQ is studied by varying the effect of concentration of collector, frother, dispersant and pH. Oleic acid and dodecylamine are used as the anionic and cationic collectors respectively. The parameters that gave the best results in cell flotation are utilized in 100 mm diameter flotation column. All the results are compared with regard to the iron grade and recovery in concentrate. It is observed that, anionic collector with dispersant dosage of 220 g/ton, at pH ~8.5 is selective and yielded a high-grade iron concentrate with low silica as impurity. Also, it is possible to achieve an iron concentrate with 64% Fe at 30-37% weight recovery. The characterization studies of the BHQ ore using X-ray diffraction (XRD) technique and liberation analysis with QEMSCAN® are carried out. The sample contains 58.8% hematite and 37.7% quartz and the liberation of hematite particles are found to be 90.3% in -25 µm size. An alternative statistical modeling approach called response surface methodology was also undertaken to optimize the operating parameters in order to maximize the iron grade and recovery. The predicted results by the model are in agreement with the experimental values.

Sep 1, 2012

By Lev Filippov, Vyacheslav Severov, Inna Filippova

"Reverse cationic flotation is a very efficient method of beneficiation of oxidized iron ores when you need to separate hematite and quartz. This method can be also applied to reduce silica content in the magnetite concentrates obtained by wet low-intensity magnetic separation. However, cationic flotation in this case is less effective due to the presence of Fe-Mg-Al-bearing silicates as amphiboles in magnetite ores. These silicates are concentrated in magnetic products and determine the difficulties of their separation from iron oxides when amines and starches are used as collectors and depressants, respectively. The recalculation of the results of the electron microprobe analysis of three calcic Fe-Mg-Al-bearing amphiboles used to characterize cation distributions in structural units and to obtain structural formulas in accord with stoichiometric limits was conducted. The Mössbauer spectroscopy was used to examine the validity of the empirical estimation of cation distributions in amphiboles. The studied samples are metamorphic pargasite and ferrotschermakite and volcanic magnesiohornblende. The crystallographic analyses of the amphibole samples exhibited their heterogeneous surfaces explaining worse floatability of amphiboles with dodecylamine at pH 10. The electrokinetic measurements showed that the position of the isoelectric point of amphiboles is related to a substitution of Al3+ for Si4+ in tetrahedral sites and to an amount of Mg2+ cations in octahedral sites. Thus, the dodecylamine adsorption onto amphiboles through electrostatic interactions is mainly affected by a distribution of these cations in tetrahedral and octahedral sites of amphiboles. In addition, the effect of starch on the depression of Fe-Mg-Al-bearing amphiboles during flotation can be attributed to the presence of metal ions on the amphibole surface, which are capable of forming strong chemical complexes with starch molecules."

Jan 1, 2014

Over the last four years, a major collaboration between CSIRO-MIS, UCT and JKMRC has produced two significant advances in the modelling of comminution. The Universal Comminution Model (UCM) uses parametric SPH/DEM studies of comminution and separation devices to calibrate mathematical models which are suitable for industrial use. The Virtual Comminution Machine (VCM) uses the same computationally intensive techniques based on discrete element modelling (DEM) and smoothed particle hydrodynamics (SPH) to directly model any comminution device by combining a CAD file describing the machine design, sophisticated ore characterisation methods and advanced particle scale flow and breakage simulation. The UCM should be useful for process analysis and optimisation of existing equipment and circuits while the VCM should make a major contribution to the evaluation and development of new machines. Progress to date is encouraging with several processes now well described using DEM. This paper also reports on progress towards combining DEM for particles with a realistic slurry model based on SPH for different types of comminution devices. In due course, this combination should allow accurate modelling of particle transport in wet mills to be combined with DEM models.

Jan 1, 2014

"The first 90 years of the last century generated a considerable volume of technical literature on the subject of grinding mill power and its prediction. It is therefore surprising that during this time there were no published models for predicting grinding mill power draw which were validated using a wide range of comprehensive industrial scale data.Notwithstanding this lack of data, the majority of models placed limited emphasis on the internal dynamics of mills and relied on simple assumptions which considered the charge to take up a fixed position and shape.Laboratory based studies (Liddell, 1986), and industrial data (Morrell, 1993), showed that these assumptions did not hold over a wide range of operating conditions, and brought into doubt the ability of the (then) existing models to accurately predict grinding mill power draw. Furthermore no model treated AG,SAG and ball mills as a single class of devices whose power draw could be predicted with the same equations.Through a detailed investigation of the charge dynamics, coupled with a large data base of industrial mill power draws, this situation was rectified by the end of the century, by which time a model had been developed and validated, which accurately predicted the power draw of all wet tumbling mills. This paper describes this model and reviews some of the earlier attempts at mill power modelling."

Jan 1, 2003

"The first 90 years of the last century generated a considerable volume of technical literature on the subject of grinding mill power and its prediction. It is therefore surprising that during this time there were no published models for predicting grinding mill power draw which were validated using a wide range of comprehensive industrial scale data.Notwithstanding this lack of data, the majority of models placed limited emphasis on the internal dynamics of mills and relied on simple assumptions which considered the charge to take up a fixed position and shape.Laboratory based studies (Liddell, 1986), and industrial data (Morrell, 1993), showed that these assumptions did not hold over a wide range of operating conditions, and brought into doubt the ability of the (then) existing models to accurately predict grinding mill power draw. Furthermore no model treated AG,SAG and ball mills as a single class of devices whose power draw could be predicted with the same equations.Through a detailed investigation of the charge dynamics, coupled with a large data base of industrial mill power draws, this situation was rectified by the end of the century, by which time a model had been developed and validated, which accurately predicted the power draw of all wet tumbling mills. This paper describes this model and reviews some of the earlier attempts at mill power modelling.INTRODUCTIONThe early 1900?s saw the beginnings of published papers which tackled the problem of what influenced the power draw of tumbling mills and what equations (models) were suitable for predicting power draw. This interest continued for the rest of the 20th century and the literature contains many papers that attempted to satisfactorily answer these questions. A common and disappointing theme in the majority of them, however, is the distinct lack of suitable experimentally derived power draw data. This deficiency consequently limited the practical application of many of the attempts at modelling mill power draw. Prior to the early 1990?s some limited data on the power draw of grinding mills in the literature can be found, but as Harris et al (1985) remarked they are ?... too frequently unusable simply because one or more essential variables have been omitted?. The absence of published results, from vigorous experimental testing of the various models, has resulted in a general lack of evaluation of the validity of assumptions and hypotheses that such models contain. Harris et al (1985) however, did make some efforts in attempting to evaluate the performance of various power prediction equations and in so doing further developed a semi-empirical one of their own. However, in using manufacturers? published data, which themselves were undoubtedly generated by proprietary equations of unproven validity, they probably succeeded in generating little more than a consensus of the various errors in these models."

Jan 1, 2003