Search Documents

By Ming-Feng Ye, Guang-liang Wu

The microstructure and mechanical strength variations of chromite pellets preheated and roasted in air were studied systematically. The results show that the chromite (Fe,Mg)(Cr,Fe,Al)2O4 is oxidized to magnesium-rich (Mg,Fe)(Cr,Fe, Al)2O4 and ferrochromium complex oxide (Cr,Fe,Al)2O3 in the preheating process over 500 °C. When the temperature is higher than 1000 °C, the new phase (Cr,Fe, Al)2O3 is formed, which is mainly distributed in the outer layer of a single chromite particle. Inner the particles there are acicular (Cr,Fe,Al)2O3 and unoxidized (Fe,Mg) (Cr,Fe,Al)2O4, which forming an interwoven structure together. The oxidation rates of pellets can reach to superior limits in 20 min. With the temperature increasing from 950 to 1300 °C, the mechanical strength and oxidized rates increase to varying degrees correspondingly.

Mar 1, 2018

By H. D. Jacobs

The Bureau of Mines undertook a research program to insure an adequate supply of iron raw materials for future iron and steel making needs. A sample was obtained of oxidized taconite from the western Mesabi Range in Minnesota, and the metallurgical response of the taconite to beneficiation by a process that included selective flocculation-desliming-cationic flotation was evaluated. The sample contained 35.3 pct iron and represented approximately 2 billion tons of crude material. Selective flocculation and flotation were evaluated in a pilot plant with a feed rate of 900 Ib/hr. Concentrates averaging 61.7 pet iron and 5.1 pet silica were obtained, with an accompanying iron recovery of 67.2 pet. About 83 pct of the process water requirements were reclaimed from the slime and froth tailing and then recycled.

Jan 1, 1980

By Herbert Clark Hoover, Lou Henry Hoover

As stated in the preface, the nomenclature to be adopted for weights and measures has presented great difficulty. Agricola uses, throughout, the Roman and the Romanized Greek scales, but in many cases he uses these terms merely as lingual equivalents for the .German quantities of his day. Moreover the classic language sometimes failed him, whereupon he coined new Latin terms adapted from the Roman scale, and thus added further confusion. We can, perhaps, make the matter dearer by an illustration of a case in weights. The Roman centumpondium, composed of zoo librae, the old German ceniner of 100 pfundt, and the English hundredweight of 1112 pounds can be called lingual equivalents. The first weighs about 494,600 Troy grains, the second 721,900, and the third 784,000. While the divisions of the centumpondium and the centner are the same, the Libra is divided into 12 unciae and the pfundt into 16 unizen, and in most places a summation of the units given proves that the author had in mind the Roman ratios. However, on p. 509 he makes the direct statement that the centumpondium weighs 146 librae, which would be about the correct weight if the centumpondium referred to was a ceniner. If we take an example such as " each centumpondium of lead contains one uncia of silver", and reduce it according to purely lingual equivalents, we should find that it runs 24.3 Troy ounces per short ton, on the basis of Roman values, and 18.25 ounces per short ton, on the basis of old German. If we were to translate these into English lingual equivalents of one ounce per hundredweight, then the value would be 17.9 ounces per short ton. [ ]

Jan 1, 1950

By K. R. Van Horn, L. W. Kempf

Plastic deformation of most commercia1 metals within a sufficiently low temperature range results in profound changes in structure and properties, of which the causes and effects are not completely understood. One of these effects is the presence of internal residual stresses and strains, which ordinarily do not appear to require specific consideration in most structural applications but which may be quite undesirable in particular instances. For example, in heavy sections that require considerable machining before assembly, residual stresses may be undesirable because of the distortion they cause when their balance is disturbed by machining or otherwise. In addition, the distribution of stresses may be such that it acts in the same direction as the applied load, thus bringing about premature failure or reducing the load-carrying capacity of the member. One source of residual stress in metal structures is the plastic deformation that may occur during rapid cooling from relatively elevated temperatures, as in the cooling of a casting in the mold or the quenching of a heavy section from elevated temperatures during heat-treatment. These residual stresses are usually undesirable because their direction of operation and magnitude cannot be controlled well enough to permit utilization of the changes in properties of the metal that they may bring about. The most convenient manner of controlling such residual stress is by proper selection of the cooling rate or by reheating to a temperature at which plastic flow can take place, resulting in relief of the residual stresses and strains. Both of these methods of control, however, may cause a reduction in the optimum properties obtainable from the previous operation. A compromise must be struck, and if it is to result in the best possible combinations of properties, specific information must be available regarding the magnitude of the stresses and the rate at which they are affected by variations in cooling rate or in reheating temperature. The complete quantitative estimation of residual stress in complicated metal shapes is exceedingly difficult, if not impossible. In simple sections, however, relatively quantitative analysis may be applied and useful cstimations of stress distribution may be made. This paper is concerned largely with the rate of relief, upon reheating, of internal residual stress in simple aluminum-alloy structures. Since even with simple structures quantitative estimation of residual stresses is laborious and costly, an attempt was made to find other means of collecting data upon which commercial heat-treatments might be based. Incidentally, a few data were collected bearing on the problem of relaxation of aluminum alloys. Determination of Residual Stresses The magnitude and distribution of residual stresses in simple aluminum-alloy shapes rapidly cooled from elevated temperatures has been discussed in a previous paper.' Fig. I reproduces some of the data

Jan 1, 1942

By L. W. Kempf, K. R. Van Horn

Plastic deformation of most commercia1 metals within a sufficiently low temperature range results in profound changes in structure and properties, of which the causes and effects are not completely understood. One of these effects is the presence of internal residual stresses and strains, which ordinarily do not appear to require specific consideration in most structural applications but which may be quite undesirable in particular instances. For example, in heavy sections that require considerable machining before assembly, residual stresses may be undesirable because of the distortion they cause when their balance is disturbed by machining or otherwise. In addition, the distribution of stresses may be such that it acts in the same direction as the applied load, thus bringing about premature failure or reducing the load-carrying capacity of the member. One source of residual stress in metal structures is the plastic deformation that may occur during rapid cooling from relatively elevated temperatures, as in the cooling of a casting in the mold or the quenching of a heavy section from elevated temperatures during heat-treatment. These residual stresses are usually undesirable because their direction of operation and magnitude cannot be controlled well enough to permit utilization of the changes in properties of the metal that they may bring about. The most convenient manner of controlling such residual stress is by proper selection of the cooling rate or by reheating to a temperature at which plastic flow can take place, resulting in relief of the residual stresses and strains. Both of these methods of control, however, may cause a reduction in the optimum properties obtainable from the previous operation. A compromise must be struck, and if it is to result in the best possible combinations of properties, specific information must be available regarding the magnitude of the stresses and the rate at which they are affected by variations in cooling rate or in reheating temperature. The complete quantitative estimation of residual stress in complicated metal shapes is exceedingly difficult, if not impossible. In simple sections, however, relatively quantitative analysis may be applied and useful cstimations of stress distribution may be made. This paper is concerned largely with the rate of relief, upon reheating, of internal residual stress in simple aluminum-alloy structures. Since even with simple structures quantitative estimation of residual stresses is laborious and costly, an attempt was made to find other means of collecting data upon which commercial heat-treatments might be based. Incidentally, a few data were collected bearing on the problem of relaxation of aluminum alloys. Determination of Residual Stresses The magnitude and distribution of residual stresses in simple aluminum-alloy shapes rapidly cooled from elevated temperatures has been discussed in a previous paper.' Fig. I reproduces some of the data

Jan 1, 1942

H. H. Kellogg (Columbia University)—The accurate measurements of the equilibrium gas ratios for the reaction: Pb(1) + H2S = PbS(c) + H2 [I] reported in this paper are a distinct contribution to the literature, and the authors are to be congratulated. I wish to comment, however, on the correlation of the data which is presented. The authors show a straight line correlation of log H2S/H2 vs 1/T, Fig. 1, and are satisfied with this for calculation of ?Fo even when extrapolated to low temperature (427 °C). Heat capacity data for the constituents of this reaction are available (at least to 900°K), however, and ACP for reaction [I] is finite and positive in the temperature range studied. Simplification of the heat capacity equations yields: ?CP= 2.71 + 1.65 x 10-3T (298°-900°K) This positive value of ?CP should result in a small but noticeable curvature of a plot of log K vs 1/T. That the experimental data fail to show such a curvature may be traced to another oversimplification of the author's treatment of their data. They assume that the activity of liquid lead is unity for all temperatures, so that

Jan 1, 1961

By A. Kenneth Schellinger, O. Cutler Shepard

Overgrinding, Or the breaking of ore particles into sizes smaller than required for liberation, is a first-magnitude problem in grinding for concentration processes. The conventional ball mill-classifier circuit used in commercial grinding produces mineral particles that are all ground to pass an upper size limit, while the proportion of very fine sizes, or slimes, is uncon rolled. Calculations based on measurements made in commercial plants which were grinding ore for flotation concentration indicate that about go pct of the energy used in grinding is consumed in useless overgrinding.1 In addition to accounting for the consumption of a large amount of energy, over-grinding is detrimental to the flotation process. Overground slime coats larger partitles and inhibits selective action by the flotation reagents on particles which by themselves could be separated very well. A further difficulty is that extremely finely ground minerals do not respond satisfactorily to flotation-collecting reagents, and both selection and recovery decrease markedly with particles finer than 5 to 10 microns.2 conventional classifier-ball mill circuits inherently tend to overgrind the valuable minerals more than the gangue minerals because most valuable minerals are of higher density and are softer and more friable than the gangue minerals. The rem- edy to this situation would seem to be a separating mechanism which would separate free valuable mineral particles regardless of size and density considerations and a mechanism which would operate in the grinding mill so that valuable mineral grains would be removed from the grinding machine as soon as they are freed. A consideration of these ideas led the authors to the conception of a simultaneous grinding and flotation mechanism. So far as the authors are aware, such a machine has never been tried, even on a laboratory scale. Development oF Apparatus and Experimental Method Apparatus development went through several stages before a really workable machine was devised. Work was started with a large cast-iron mortar 7½ in. in diameter and 8 in- deep. A cast-iron disk Was mounted on the end of a vertical shaft SO that it fitted loosely into the bottom of the mortar. On top of the disk four radial cleats ¼ in. high were bolted to agitate the ball charge as the disk, or impeller, rotated. The vertical shaft was driven at 100 rpm by a mechanism improvised from an old coffee-mill type of grinder. A hole was drilled into the bottom of the mortar through which compressed air was introduced for frothing. In operation, more charges of Ore which had been crushed to pass an 8-mesh screen were placed in the bowl along with water and suitable reagents. The impeller was then started and the compressed air turned On so that the ore was ground while being exposed to air bubbles. Mineralized froth was

Jan 1, 1949

By F. W. Wessel

This report summarizes Bureau of Mines laboratory tests of four tungsten¬-bearing materials of various grades, The test work has demonstrated that the dry reaction between sodium carbonate and scheelite is best conducted at 900° C. or a few degrees less. The sodium carbonate requirement, in addition to the stoichiometric equivalent of the scheelite, is in some inverse relationship to the grade of tungsten in the ore. Residence time and particle size exert a definite but minor control on the reaction. Pressure-digestion tests on an impure concentrate indicate the technical feasibility of extraction by this process. INTRDUCTION Over 100 years ago Oxland discovered that tungstic oxide could be removed from wolframite and converted to a soluble form by reacting the mineral with a soluble alkaline salt at elevated temperatures. He patented this discovery, and this patent 4/ since has been the basis for most tungsten extraction. Much literature exists on this process and its modifications, and the characteristics and possibilities of the reaction are quite well defined.

Jan 1, 1956

By Michael Sokaski

A dense-medium cyclone pilot plant was built to clean 112-inch to 0 size coal. This pilot plant does not require the removal of the extremely fine material (usually minus 1/2-mm) from the feed, which is necessary in conventional dense-medium cyclone plants. The essential difference between conventional cyclone plants and the pilot plant is in the medium-recovery system. The latter uses a combination of sieve bend screens, classifying cones, and magnetic separators for medium recovery. Eight tests were run in the pilot plant at a feed rate of 5 tons per hour with four different raw coals and one crushed jig middling product. Sharp separations were obtained as indicated by an average error area of 27 and probable error of 0.047 for the 112-inch to 200-mesh size range. Magnetite loss averaged about 2.5 pounds per ton of feed. The influence of several factors on cleaning performance of the cyclone was investigated. The diameters of both the overflow and underflow orifices were found to be important for optimum cleaning performance, but these diameters could vary over a wide range without serious effect; however, if this range was exceeded, cleaning performance deteriorated. Better cleaning performance was obtained when the cyclone inlet pressure was increased from 9 psig to 13 psig. The size of the magnetite particles in the medium was found to have a substantial effect, and the best cleaning results were obtained with grade B magnetite. Demagnetizing the magnetite in an effort to decrease the effective particle size did not improve cyclone performance.

Jan 1, 1963

By J. B. Rubinstein

"The paper gives an overview of the history of flotation machine development. The laws of their evolution are stated and the place and role of column flotation are demonstrated.The major parameters of column cell diameter and height are discussed. Options are proposed for optimization of the hydrodynamic characteristics of large-capacity column cells. The high efficiency of columns with cell height of 4-6 m is demonstrated. The optimum parameters of multisectional column machines are determined. Their advantages compared to monocells are shown. A scaling-up procedure for column machines was developed based on flotation macro- and microhydrodynamics simulation."

Jan 1, 2003

By Collin L. Stewart

A summary of the results of a limited subsidence monitoring program at Cyprus Amax Coal Company's Shoshone Mine No. 1 in the Hanna Basin of south central Wyoming is presented. The monitoring program was designed to meet regulatory requirements for confirmation of subsidence predictions made in the mining permit. The longwall panels were 380 ft to 600-ft wide. Mining heights varied from 11 ft to 13.5 ft. Gate road development included three-entry and two-entry head and tail gates. Overburden depth varied from 150 to 945 ft. The coal seam and overburden dip at approximately 11 degrees. The overburden is very low to low strength mudstone and carbonaceous shale with low to moderate strength, silty sandstone interbeds. Panel W/h ratios varied from subcritical to supercritical. Smax was 0.9 of the mining height. The subsidence ratio was about 0.6 for subcritical width. The draw angle on the up-dip side averaged 20 deg. The draw angle on the down-dip side averaged 12 deg.

Jan 1, 1995

By W. S. Dyer

Introduction In northern Ontario, refractory clay is found on four rivers of the James Bay watershed: the Abitibi, the Mattagami, the Missinaibi, and the Moose. The clay all belongs to the same geological formation ( the Mattagami of Lower Cretaceous age) , and occurs within a structural depression or 'basin', the boundaries of which have never been defined but which appears to cover at least 1,500 square miles of territory. The recently constructed James Bay extension of the Temiskaming and Northern Ontario railway crosses the northeast end of the basin at the Onakawana lignite field, which by rail is 126 miles north of Cochrane, 605 miles north of Toronto, and 60 miles southwest of Moose Factory. Smoky Falls, the northern terminus of the railway owned by the Spruce Falls Power & Paper Company, Limited, is 7 miles south of one other point in the basin, i.e., the outcrops of refractory clay, sand, and kaolin at the foot of Long rapids on the Mattagami river. Smoky Falls is 50 miles north of Kapuskasing. In a very early report(l) , Dr. Robert Bell, of the Geological Survey of Canada, described '"light coloured marl in horizontal beds" from the Missinaibi river, and noted that the lignite also occurring in the river was probably :associated with the marls. This was undoubtedly the Cretaceous refractory day. Other early writers noted the occurrence of fire-clays, but it was not until Joseph Keele(2), of the Geological Survey, visited thee region in 1919 and studied the clays that their real significance was known.

Jan 1, 1933

Techniques and results achieved during the excavation of the Melbourne Underground Rail Loop (HURL) are summarised and discussed in the context of the various restraints present in a typical urban construction project. Mechanical methods (as distinct from those utilising blasting)accounted for the bulk of the excavation. Ground support was critical in this particular geological excavation setting, but is not discussed in detail except insofar as it constitutes a restraint to excavation progress.

Jan 1, 1978

By J. Kabuya Mukendi, F. Brunet, J. Hadjigeorgiou, P. Matte, D. Leblanc, M. Grenon

La stabilité des bancs miniers est un élément important de la conception des fosses. Il n’y a pas de consensus sur les critères à utiliser pour évaluer la fiabilité des bancs miniers. Lors de la conception des fosses, la fiabilité des bancs miniers est analysée à l’aide d’un ou deux critères. L’objectif de cet article est d’analyser l’efficacité de la conception des bancs miniers de la fosse Tiriganiaq du projet minier Meliadine par rapport à des critères de fiabilité rencontrés dans la littérature et la pratique. L’article présente les similitudes et les différences entre les résultats obtenus pour les différents critères et propose une approche multicritère systématique pour évaluer la fiabilité des bancs miniers à l’aide de cartes de susceptibilité.

Jan 1, 2012

By D. A. Reynolds, R. G. Travis, James J. Dowd, R. F. Abernethy

"1. Reserve estimates have been made for eight coal beds in Monongalia County, W.Va., but only the Pittsburgh, Sewickley, and Upper Freeport beds are important at this time. The Pittsburgh, Sewickley, Waynesburg, and Upper Freeport beds contain the largest recoverable reserves.2. Known measured and indicated reserves in all beds, based on a minimum thickness of 14 inches and 1,800 short tons per acre-foot of coal in place, are estimated at 3,526 million tons as of January 1, 1954. Of this total, 3,294 mil¬lion tons is in beds 28 inches and more thick. Areas in each bed except the Pittsburgh were omitted from the estimate because available data relative to the bed characteristics are too meager for making an estimate that conforms with the definitions of measured and indicated coal adopted for this study. Should future drilling or development prove reserves in these areas, such reserves should be added to the total estimated reserves.3. Recoverable reserves of coal have been estimated in beds 28 inches and more thick. This thickness is about the minimum now being mined by hand loading onto conveyors in the Appalachian region. The weighted average recovery for all beds in Monongalia County, as determined by this investigation, is 47.9 percent. This percentage is based on the total thickness of coal (less partings three-eighths inch thick or more) in the bed rather than on the thickness of the coal mined. Based on the weighted average percentage of recovery for all beds in Monongalia County, the recoverable reserves are estimated at 1,577 million tons as of January 1, 1954.4. Monongalia County coals of the major producing beds are of high-volatile a bituminous rank.5. The Pittsburgh and Lower Freeport beds in this county yield coal that can be used as metallurgical coal, especially when blended with low-volatile coals. The Lower Freeport coal is coked singly and in beehive ovens, and makes good metallurgical coke.INTRODUCTIONThe investigation to evaluate the reserves of coking coal was planned to cover three phases: (1) Estimation of known measured and indicated recoverable reserves of all coking coal; (2) study of methods of upgrading marginal coals through effective preparation; and (3) study of carbonizing properties of coals and coal blends not now widely used for metallurgical coke making."

Jun 1, 1956

By P. S. Kotval, O. H. Nestor

Stacking fault energy measurements of nickel-base alloys have been mainly confined to binary and ternary systems. In this paper, the stacking fault energy has been measured by the rolling texture method in a series of ten alloys which comprise successive additions of Cr, Mo, Fe, and C to pure nickel, eventually resulting in an alloy of the composition of Hastelloy alloy X. The alloys studied here are single-phase, solid solutions with the exception of two alloys in which some undissolved particles of "primary" carbide have been retained. It is found that successive additions of chromium, molybdenum, and iron all lower the stacking fault energy, with iron having only a minor effect. The stacking fault energy is found to increase when carbon is added in solid solution. The results from the rolling texture measurements are correlated with thin foil observations of dislocation substructures in these alloys. In a recent paper' it was pointed out that the dislocation substructure of various superalloy matrices could be classified into three broad categories based on 'high', 'medium', and 'low' stacking fault energy. It has also been demonstrated2 that the dislocation substructure in each of these categories has a well defined role in the nucleation of strengthening precipitates which is different from the role played by the dislocation substructure in other categories. Thus, it becomes desirable to understand the influence of various solute elements on the stacking fault energy and hence on the dislocation substructure of the matrix, before any further development of superalloys by mi-crostructural predesign can be undertaken. Recently, Beeston and France have studied the influence of increasing solute additions on the stacking fault energy of a series of binary nickel-base alloys relevant to the Nimonic series using the rolling texture method, and have then estimated the effect of a given alloy addition in five commercial Nimonic alloys. However, comparison with stacking fault energy data from other investigations''5 suggests that the influence of a given solute element in a nickel-base binary system is not necessarily the same in a ternary or more complex superalloy system. Accordingly, the present work was undertaken to study the effect of successive addition of solute elements to pure nickel, the final composition being the nominal composition of Hastelloy X. The rolling texture method of stacking fault energy measurement was used since it can be used for the whole range of stacking fault energy values and does not have the disadvantage of, say, the Node method which is only applicable to low values of stacking fault energy. In addition, the rolling texture method provides a means of determining the stacking fault energy which is statistically more significant than that provided by other methods. EXPERIMENTAL TECHNIQUES Button heats of alloys of the compositions shown in Table I were prepared. Each button was remelted not less than four times. After a slight deformation (approximately 5 pct) all alloys were homogenized at 2200°F except alloys, H . I, and J. Alloys H and I were solution heat treated at 2150°F and alloy J at 2282OF. The buttons were cold worked by rolling, using "end-to-end" passes and intermediate anneals at the homogenization temperatures mentioned above. After each annealing treatment the samples were rapidly water quenched to avoid any precipitation. In alloys F and I, however, a few particles of "primary" carbides were retained even after the homogenization treatments at the temperatures mentioned above. Part of the solution heat treated material was cold worked to 0.04-in.-thick sheet and the penultimate reduction was -50 pct of deformation as recommended by Dillamore et al. All annealing was carried out in vacuo within sealed quartz capsules. Some of the material from each alloy was rolled down further to 0.004 in. strip for thin foil transmission electron microscopy specimens. Specimens of this strip were annealed at the homogenization temperature for 1 hr and then strained 7 pct by rolling at room temperature. Thin foils were prepared from the strip specimens by the 'window" technique using an Ethanol-Perchloric acid electrolyte at 32°F and a voltage of 22 v. Stainless steel cathodes were employed. All transmission electron microscopy was performed in a JEM-7 electron microscope using an accelerating voltage of 100 kv. Specimens from the 0.04 in. sheet which had been rolled -60 pct in the final pass were electropolished to remove the surface layers to a depth of approximately 0.002 in. Rolling texture pole figures for all the alloys were determined using a Schulz ring and nickel filtered CuKa radiation at 50 kv and 20 ma. The texture parameter Io/(lo + I,,) (where Io is the

Jan 1, 1970

By C. V. O. Hughes

DIATOMACEOUS earth is probably the most widely used, and certainly the least well publicized, of the important non-metallics of the United States. The very multiplicity of the names by which this whitish earthy sediment is described has served to confuse its identity, and to dissipate a realization of its full importance.

Jan 3, 1953

By Wiston Rocher

The Minera Ojos Del Salado (MINOSAL) concentrator is located 20 km (12 miles) southeast of Copiapo, Chile. It processes copper ore provided by its Santos and Resguardo mines at a capacity of 4.2 kt/d (4620 stpd). In 1982, Phelps Dodge acquired the majority interest in MINOSAL to explore the property's potential. At that time, the concentrator plant was a small, antiquated installation with a capacity of 674 t/d (740 stpd). At the time of the pur¬chase, the majority of the ore was produced from the Santos and Resguardo mines, but custom ores of copper, silver and gold were also treated. The acquisition occurred at a crucial time in the copper business, since the price had fallen to nearly 50 cents/ lb. In 1985, Phelps Dodge purchased the outstanding interest in MINOSAL and acquired 100% of the property. Today, after numerous improvements and two expansions, the concentrator is a combination of old and new equipment. The regrind mills, for example, have been in service for about 50 years, while the ceramic concentrate filter represents the latest technology.

Jan 1, 1993

The O.K. deposit occurs within a belt of cherts and spilitised basic volcanics near the western margin of the Hodgkinson Province. The tabular to lenticular shape and gross conformability of the ore bodies; the common occurrence of colloform, framboidal and banded textures in the ores; the lack of significant alteration adjacent to the massive ore horizons and the apparent restriction of stockwork mineralization to the foot-wall side of the massive ore zone at the O.K. mine, all favour a syngenetic origin for the massive sulfide bodies. Furthermore, the micro- structural evidence suggests that the ores have undergone the same deformational history as the country rocks. The invariable association of the ores with spilites; the S isotope values and Co-Ni contents of the pyrites, together with the primitive metal assemblage (Fe, Cu, minor Zn) strongly favour a basic magmatic source for the sulfides. It is suggested-that the sulfides were probably precipitated sub aqueous- ly during a hiatus in volcanic activity. The O.K. deposit is of the cupreous pyrite variety, but unlike other well known occurrences of this type, it does not con- stitute part of an ophiolite complex. While the host volcanics exhibit strong ocean floor tholeiitic affinities, no associated ultra

Jan 1, 1978

By A. Kenneth Schellinger, O. Cutler Shepard

INTRODUCTION OVERGRINDING, or the breaking of ore particles into sizes smaller than required for liberation, is a first-magnitude problem in grinding for concentration processes. The conventional ball mill-classifier circuit used in commercial grinding produces mineral particles that are all ground to pass an upper size limit, while the proportion of very fine sizes, or slimes, is uncontrolled. Calculations based on measurements made in commercial plants which were grinding ore for flotation concentration indicate that about 90 pct of the energy used in grinding is consumed in useless overgrinding.1 In addition to accounting for the consumption of a large amount of energy, overgrinding is detrimental to the flotation process. Overground slime coats larger particles and inhibits selective action by the flotation reagents on particles which by themselves could be separated very well. A further difficulty is that extremely finely ground minerals do not respond satisfactorily to flotation-collecting reagents, and both selection and recovery decrease markedly with particles finer than 5 to 10 microns.2 Conventional classifier-ball mill circuits inherently tend to overgrind the valuable minerals more than the gangue minerals because most valuable minerals are of higher density and are softer and more friable than the gangue minerals. The remedy to this situation would seem to be a separating mechanism which would separate free valuable mineral particles regardless of size and density considerations and a mechanism which would operate in the grinding mill so that valuable mineral grains would be removed from the grinding machine as soon as they are freed. A consideration of these ideas led the authors to the conception of a simultaneous grinding and flotation mechanism. So far as the authors are aware, such a machine has never been tried, even on a laboratory scale. DEVELOPMENT OF APPARATUS AND EXPERIMENTAL METHOD Apparatus development went through several stages before a really workable machine was devised. Work was started with a large cast-iron mortar 7 ½ in. in diameter and 8 in. deep. A cast-iron disk was mounted on the end of a vertical shaft so that it fitted loosely into the bottom of the mortar. On top of the disk four radial cleats ¼ in. high were bolted to agitate the ball charge as the disk, or impeller, rotated. The vertical shaft was driven at 100 rpm by a mechanism improvised from an old coffeemill type of grinder. A hole was drilled into the bottom of the mortar through which compressed air was introduced for frothing. In operation, weighed charges of ore which had been crushed to pass an 8-mesh screen were placed in the bowl along with water and suitable reagents. The impeller was then started and the compressed air turned on so that the ore was ground while being exposed to air bubbles. Mineralized froth was

Jan 1, 1947