Search Documents

By J. E. Gill, A. R. Byers

"The property of Sladen Malartic Mines, Limited, is in the northern half of Fourniere township, Abitibi county, Quebec. The group is large, but the part so far found to be of economic interest is a strip 3,200 feet long between the Canadian Malartic property to the west and the East Malartic and National Malartic properties to the east ( Fig. 1).The original claims were staked in 1923 and 1924 by St. B. Sladen and H. S. Kennedy because gold was known to occur on claims nearby. In 1928, detailed mapping of the Malartic Gold Mines property (now Canadian Malartic) showed the presence of an east-west gold-bearing zone traceable almost to the Sladen boundary. During the winter of 1929, a hand-driven drill was used to secure short sections of core through casing pipe sunk to bed-rock. This proved the existence of the main Sladen ore zone, but it was not until the price of gold was increased, in 1934, that serious work was attempted.Underground work was started in 1935 and milling in January, 1938. To the end of 1945, 168,882 oz. of gold were recovered from 1,641,974 tons of ore, an average of 0.103 oz. a ton.GEOLOGYThe bed-rock underlying' the Sladen Malartic property consists mainly of Precambrian greywacke and volcanics that appear to form a conformable series with a general northwest strike and vertical to steep northerly dip. Except where disturbed locally by drag folds or faults, tops face north. Intrusive rocks that range in composition from diorite to granite cut the sediments and volcanics. The basic types, generally quartz diorite and diorite, form dykes or sills in the volcanic rocks. The more acidic types, mainly quartz syenite and related porphyries, intrude the volcanics and the sediments in the form of dykes, sills, and larger, stock-like bodies.The orebodies are parts of slightly shattered or brecciated zones, mainly in greywacke, but partly in syenite, diorite, and volcanics, that have been more or less replaced by fine-grained quartz, pyrite, and minor amounts of sphalerite, galena, sylvanite, petzite, and gold."

Jan 1, 1954

By Julie Hunt

Rimfire Minerals Corporation: early stage precious metal exploration specialists with discovery opportunities in partnership with major and mid-tier gold producers or prominent junior exploration firms. Mike Roberts, Senior Geologist

May 1, 2006

By Martin Bergeron

Producteur d?or de taille intermédiaire Mines Aurizon Ltée Bureau-chef àVancouverMine et projets localisés au Québec2

May 1, 2011

By Jonathan Peck, Jim Gray

"With the implementation of more advanced monitoring, control and locating systems in surface mines, there is an ever increasing need for their mine-wide integration toward achieving over-all optimization. AQUILA Mining Systems Ltd. is developing the basis for a comprehensive real-time monitoring, control and information system called the Total Mining System or TMS. This system is being designed to address both current and future needs of open pit mines toward improving productivity and minimizing costs, through better two-way information flow between each component in the operation. At a basic level, feedback from equipment and process monitoring will enable decision making to be based on quantitative, on-demand information, thus permitting proactive rather than reactive planning. Dynamic planning would be a valuable tool as market conditions fluctuate, equipment fleets age and variations in the grade, quality, geometry and hardness of the mined material all contribute to either changing production targets or levels of achievable performance. Such a capability could be implemented within the next two years.At a higher level, and within five years, autonomous operation and/or real-time supervisory control of mobile and stationary equipment in open pit mines will be achieved. A discussion is provided in regard to the development and implementation of an autonomous haultruck and blasthole drill, to demonstrate the need for such a system.This paper outlines a preliminary program for the design, development and implementation of this system, primarily in relation to high-level communications and its role in the future automation of open pit mobile equipment."

Jan 1, 1995

By Matthew Russell, Bobby Randhawa, Leopoldo Mariscal Samaniego, Jose A. Puente, D. Brosig, Paykan Safe

The La Caridad copper smelter, located in Nacozari, Mexico, have started a multi-year plan to reduce sulfur dioxide and particulate matter emissions into the environment. In 2014 the smelter completed a voluntary plant-wide SO2 process emissions inventory as part of the smelter’s proactive environmental stewardship policy. In turn, this triggered the development of several environmental projects to alleviate these emissions. A strategy consisting of improved capture efficiency of the sources with greatest impact has been followed to maximize cost-effectiveness. The reduction of secondary emissions from the Teniente Converter, as well as of primary and secondary emissions from the Peirce-Smith Converters have been key areas of priority. Additionally, a new system to capture and recover particulate matter from the Anode Refining Furnaces off-gas is also in the process of being implemented. This paper presents an overview of the work developed behind this effort and the smelter’s vision towards a greener future. INTRODUCTION The La Caridad copper smelter is owned and operated by Metalurgica de Cobre S.A. de C.V. (a subsidiary of Grupo Mexico). Currently, the smelter has a production capacity of 300,000 tonnes of anodic copper per year. The smelter is coupled with a 300,000 t/y electrolytic refinery. Other by-products (such as precious metals and sulfuric acid) are also generated as part of an integrated process. The original smelter started operating in 1986 with one Furukawa concentrate dryer, one Outokumpu Flash Smelting Furnace (FSF), two slag cleaning electric furnaces (SCFs), three Peirce-Smith Converters (PSCs), two Anode Refining Furnaces (ARFs) and one casting wheel. In 1997, a plant expansion project was concluded with the installation of one steam dryer, one Modified Teniente Converter (MTC), one additional ARF, one additional casting wheel, and the electrolytic refinery [1]. Over the years, the smelter management has implemented numerous environmental projects to alleviate SO2 and particulate matter (PM) emissions. The smelter process gas handling system captures the process gases from the FSF, MTC, and the three PSCs. The process gases are cleaned and conditioned and sent to the two acid plants. The first acid plant (with a capacity of 226,000 Nm3/h, wet basis) was commissioned in 1988 to process the FSF and PSC off-gases. A second acid plant (with a capacity of 195,000 Nm3/h, wet basis) was commissioned as part of the smelter expansion project [1]. The PSCs also have a secondary gas handling system with a dedicated baghouse and stack. The Furukawa dryer off-gas reports to a dedicated electrostatic precipitator (ESP) for dedusting before venting via a local stack. Process gas from the two SCFs, along with hygiene ventilation from various sources inside the smelter (including matte and slag tapping hoods from the FSF) report to a baghouse-based gas handling system and vent to atmosphere via a local stack. The original ARFs do not have an off-gas handling system, however, each ARF has a post- combustion chamber which naturally vents the process gases to the atmosphere via a dedicated stack.

Jan 1, 2019

By C. Bazin, D. Hodouin, M. Ounpuu

Sensors are now available in mineral processing circuits to monitor pulp flowrate and density, as well as ore characteristics and other process variables. The amount of information provided by these instruments is considerable but in many cases poorly exploited. This paper gives the results of the application of time series analysis techniques to signals generated by the sensors installed on an industrial Pb dezincing flotation circuit in New-Brunswick. Autocorrelation and cross correlation functions appear to be interesting tools for a preliminary circuit analysis, however the calibration of single input single output discrete transfer functions using collected sensors signals leads to conflicting results. For this reason a phenomenological approach, based on the kinetic and transport properties of the flotation process, is also considered. It has several advantages over the empirical transfer function modelling but also limitations in its applicability to industrial processes. A combination of the transfer function and kinetic modelling approches appears to be a promising avenue for better process modelling and control.

Jan 1, 1985

By Frank Condon

?Location ?Background ?Drilling Program ?Mineral Resources ?Geology ?Economic Assessment ?Strategic Partner ?Located in Nunavik, Québec ?160 km northwest of Schefferville ?240 km south of Kuujjuaq ?160 km northeast of Brisay ?Magnetic Anomaly ?Property Boundary ?Diamond Drill Holes

May 1, 2011

By R. A. Walker, G. Scheltgen, T. Javorsky, G. Blackwell

"Recent engineering advances in open pit rock mechanics have resulted in improved pit wall slope design and improved operating procedures in the event of a major wall instability. At the other end of the spectrum, on the scale of single bench failures, considerable discretion is left to the mine operators as to how potential problems are handled. The usual alternatives are either to operate with caution beneath potential hazards, or try to bring the unstable rock down. In specific cases it may be advantageous to attempt to stabilize rock structures for reasons of access and operating safety.A t the Brenda Mines open pit a prominent structure on the north wall had caused recurring bench-scale wedge failures when exposed on each mining level. In spite of operating precautions a shovel was hit by a rockfall causing significant damage. The first section of this paper describes the investigative, design, and stabilization methods which permitted subsequent mining beneath the structure to proceed without incident. Emphasis is placed on the personnel and equipment resources available at the mine site which were used to minimize costs and which would be available at most operating mines.The second section describes an ongoing research project that utilizes radio telemetry to transmit slope monitoring data. As part of this project the wedge structure was instrumented so that it could be continuously monitored with the aid of a computer. The application demonstrated the potential of 'micro- chip' technology to remote data collection and processing as related to rock structure monitoring, slope stability and safety."

Jan 1, 1989

By N. W. D. Davis

"As I have recently carried out some further research on the Navan lead-zinc deposit data while working in Michel David's research team, I should like to add a few comments on the paper- ""Universal Kriging for Ore-Reserve Estimation- Conceptual Background and Application to the Navan Deposit"", by M. Dagbert and M. David (CIM Bulletin, pp. 80-92, February 1976). The first comment is that since this research has been carried out, important new theoretical advances have been made which make for an easier application of geostatistics to the non-stationary case (see Matheron0l). This is the theory of intrinsic random functions of order k. With the aid of this theory, it has been possible to write computer programs which can carry out automatically universal kriging, although so far such applications have been largely confined to automatic contouring and not to mining problems. However, if approached with care there is no reason why these methods should not be applied to mining problems and, in this case, an automatic study has been done on this data."

Jan 1, 1976

By Sadan Kelebek, Gerhard W. Heinrich, Peter F. Wells, Peter J. Whittaker

"Based on plant mineralogical data & circuit modelling, a pilot plant campaign was conducted at Strathcona Mill to identify options for improving the performance and operating flexibility of the copper-nickel separation circuits. Emphasis was given to options for combined treatment of the two feed streams (primary nickel ore rougher concentrate and copper ore circuit concentrate) to obtain more flexibility in the use of existing column flotation capacities. The pilot test results demonstrated that circuits which incorporated preconcentration prior to column flotation gave the best metallurgical results. Maximum column carrying rates were also related to the preconcentration. Predictions that increased circulating loads from the column scavengers would improve copper-nickel separation performance were confirmed.IntroductionThe copper-nickel separation circuits at Falconbridge's Strathcona Mill produce a high grade (>30% Cu) copper concentrate most of which is treated at the Falconbridge Kidd Creek smelter near Timmins. The remaining nickel concentrate which still contains approximately 20% of the feed copper is combined with other lower grade nickel/copper concentrates and shipped to the Falconbridge nickel smelter. The separation is required because of the limited capacity for copper processing in and downstream from the nickel smelter. Since no nickel is recovered at present in the Kidd smelter, the challenge is to minimize the nickel content of the copper concentrate.At present there are two independent separation circuits treating the rougher concentrates from the copper and nickel circuits as shown in Figure I. The minerals treated are mainly chalcopyrite and pentlandite but there is up to 25% of the copper present as cubanite in the copper ore. The circuits were kept separate because the flotation rates and froth characteristics were very different, probably because of the presence of cubanite and the large difference in the concentrate grades feeding the circuits."

Jan 1, 1995

By Andrew Artzer, Michael Moats, Jack Bender

Antimony and bismuth in copper electrorefining electrolytes can cause cathode contamination, leading to the careful control of their concentrations. This study examined the laboratory scale removal of bismuth and antimony from an electrorefining electrolyte using a solvent extraction mini-pilot plant. The ability of a proprietary phosphonic acid ester extractant (REX-2) in a commercial diluent was studied in a simulated commercial setting. Antimony and bismuth were both extracted effectively from the electrolyte (72% Bi and 70% Sb removed) and Bi was effectively stripped (92% Bi recovered) with 400 g/L sulfuric acid. A final strip solution was obtained from the SX circuit containing 0.36 g/L Bi, 0.043 g/L Sb, 1.3 g/L Cu, and 0.18 g/L As. Unfortunately, most of the extracted Sb remained in the organic phase. Following the pilot study, experiments were performed to examine the precipitation of bismuth from the aqueous strip phase and the removal of antimony loaded on to REX-2. INTRODUCTION For many years, antimony and bismuth have been some of the most problematic impurities in copper electrorefining (ER) (Bardwell & Lapee, 1933). Along with surface roughening, high concentrations of Sb and Bi can contaminate the copper cathode and precipitate at the copper grain boundaries. This drastically reduces the drawing ability of the metal (Bender & Emmerich, 2016). For these reasons, it is necessary to control Sb and Bi concentration in the cathode, which is done by controlling their concentration in the electrolyte. Conventionally, Sb and Bi concentrations are controlled by maintaining a proper As/(Sb+Bi) molar ratio in the anode and arsenic concentration in the electrolyte. Antimony and bismuth are removed during electrolyte purification using liberators (electrowinning cells) and/or ion exchange. Liberators treat an electrolyte bleed stream from the ER circuit and electrodeposit increasingly impure copper as the copper concentration of the solution decreases. Traditionally, first stage liberators reduce the copper concentration from ~45 g/L to ~15 g/L and produce saleable copper. Second stage liberators decrease the copper concentration to ~8 g/L (Biswas & Davenport, 1980) which results in contaminated cathode that is returned to anode production (Biswas & Davenport, 1980; Wang, 2004). Further electrolyte purification is not always required but third stage liberators can be used to reduce copper concentration below 2 g/L which results in significant removal of As, Sb and Bi and very impure deposits which are recycled to a smelter or used for As production. However, liberators do not have very high energy efficiency (Acharya, 2014; Ando & Tsuchida, 1997) and below 0.5 g/L copper, highly toxic arsine gas may form (Biswas & Davenport, 1980).

Jan 1, 2019

By B. McFadzean, J. Taguta

In order to recover minerals by flotation, the sub-processes of collector adsorption onto the valuable minerals followed by bubble-particle attachment should occur efficiently. This paper investigates both of these sub-processes for various thiol collectors and their mixtures onto base metal sulfide minerals. The aim of the work was to investigate whether there is a correlation between the strength of the collector interaction with the mineral and the subsequent bubble-particle attachment. The collector-mineral affinity was measured experimentally using an isothermal titration calorimeter in order to determine the molar enthalpy of adsorption for each collector-mineral system. The bubble-particle attachment was measured using a microflotation device which essentially determines the hydrophobicity of the mineral. Sodium ethyl xanthate (SEX) and sodium diethyl dithiocarbamate (DTC) and mixtures thereof were used as collectors with either single minerals or binary mixtures of pyrite and chalcopyrite, respectively. The adsorption enthalpies of the collectors were interpreted with reference to the chemical structure of the collector molecule and showed that the greater the positive inductive effect of the collector, which may be qualitatively related to their pKa, the greater their affinity for the mineral surface. In the microflotation studies, recoveries were strongly related to the alkyl chain length but there was no correlation between the enthalpies of adsorption and the flotation recoveries. When mixtures of collectors were used there was clear evidence of synergistic effects and in the case of DTC/xanthate mixtures it is possible that this synergism is related to a DTC promoted oxidation of xanthate to dixanthogen. When mixtures of pyrite and chalcopyrite were used the flotation recovery and enthalpy of adsorption of SEX onto the individual pyrite particles in the mixture increased substantially compared to the case of pyrite alone. Mechanisms are proposed to interpret these important observations.

Jan 1, 2016

By P. R. A. Andrews

"Talc ore from Bakertalc, High water, Quebec analyzing about 60% talc, previously dry ground to about 95% minus 53 µm was beneficiated in an 11.4 kg/ h pilot-plant flotation circuit. A concentrate of 95.9% talc (necessary for paper-grade) was obtained at a recovery at 75.3% by employing 53 g/t of Dowfroth 250 collector at pH 9.2; further cleaning increased the grade to 97.2% talc at a recovery of 57.5%.Flotation of fresh ground 60% minus 44 µm ore at a feed rate of 18.6 kg/ h produced a grade of 96.4% talc with a recovery of 78.2%, using Aerofroth 70 collector a six carbon-branch chain alcohol, at pH 9.2.The similarity in the concentrate grade and recovery between pre-ground and fresh-ground ore indicated that it was not neces¬sary to grind to 95% minus 53/2m to achieve comparable selectivity. A grind of 60% minus 44 pm corresponds to 100% minus 300 µm.Flotation retention times obtained with fresh ground ore indicate 33 min. for combined rougher/scavenger, and 15 min. each for the first and second cleaners.Talc-magnesite flotation separation is explained in terms of physico-chemical forces, with reference to surface charge and adsorption mechanisms."

Jan 1, 1989

By O. Bouaziz, O. Magnea, P. Cugy, E. F. Rauch

The experimental difficulty to perform Equal Channel Angular Extrusion (ECAE) of a bulk sample of hard material such as steel was overcome. The original method proposed in this paper consists in replacing the bulk sample by a stack of alternated hard metal (steel) and soft metal (aluminium) sheets. By applying this procedure to an Interstitial Free steel (IF steel), 2.7µm grain sized material was obtained after recrystallisation. Its mechanical properties (hardness, formability in shear, hardenability and faigue limit) were investigated and compared with the properties of the same IF steel with conventional grain sizes (12m and 25µm).

Jan 1, 2004

By Clifford R. Stanley

"Abstract - Reproducible analyses of representative samples for elements affected by the ‘nugget effect’ is vital in mineral exploration and reserve estimation. Unfortunately, a priori knowledge of what size of sample is representative from a mineral deposit or its dispersion train is commonly unavailable. Use of the simple ‘equant grain model’ and the NUGGET computer program facilitates determination of the sampling parameters that will produce appropriate samples for chemical analysis. Based on Poisson statistics, NUGGET can be used to determine: (1) the ‘effective grain size’ of nuggets in a suite of samples, (2) the level of representativity (precision) that samples with specific nugget grain sizes will exhibit, and (3) the sample size required to obtain a specific level of sample representativity. The program determines these sampling parameters for any element in any mineral, including solid solution minerals (e.g., Au, electrum, Pt-Fe alloy, diamond), and allows assumption of a variety of grain shapes (e.g., ellipsoids, cylinders, parallelepipeds, etc.). Furthermore, it outputs the effective number of nuggets expected in a sample of given size and the percentile confidence bounds on any element concentration affected by the nugget effect.Results from the software allow determination of effective sampling and analysis strategies that minimize the nugget effect in geochemical analysis/assaying used in mineral exploration and reserve estimation. This will not only improve confidence in the resulting exploration results or reserve estimate, but will facilitate estimation of the error associated with any subsequent interpretation or reserve figure."

Jan 1, 1998

By H. G. Ames

"The property of Perron Gold Mines, Limited, lies in the southwest corner of Pascalis township, northwestern Quebec. It is approximately 12 miles northeast of the town of Val d'Or.This ground was originally staked by Jack Mathews of Amos in the spring of 1931. Shortly after staking, a quartz vein containing visible gold was uncovered on the south boundary of the property, about 400 feet south of the present shaft. With further prospecting, another spectacular goldbearing vein was found 3,000 feet northwest of the original showing. Subsequently, several programmes of exploration were carried out in search of the main ore zone. By 1933 sufficient ore was outlined to warrant sinking a shaft. The first ore was milled in 1934.Total production from the Perron mine from commencement in 1935 to 1944 has been 317,900 oz. gold having a value of $11,935,106. The average recovery a ton for 1944 was $9.01."

Jan 1, 1954

By D. R. Hatton

To obtain sufficient material on which to perform copper-molybdenum separation testwork a variability set of drill hole reject samples amounting to several tons was put together to include the different ore types proportions expected in the copper circuit. The samples were treated in a copper circuit pilot plant and enough mass of bulk concentrate was recovered to conduct copper moly separation tests in the laboratory. Tests consisted of stage flotation reactor (SFR) rougher pilot plant tests, conventional laboratory rougher, cleaner and 4 stage open circuit cleaner tests. The mineralogical liberation analysis of the feed and concentrate was used to model the effect of liberation on the grade-recovery relationship of the different samples. Mineral kinetic parameters for the various samples were determined by calibration to the test data. The copper concentrate results as simulated from previous studies on the same deposit were used to simulate the variability in head grade and throughput expected for the feed to the copper molybdenum separation circuit for each of the first 10 years of production. This variation in feed was then simulated through the separation circuit to obtain optimised molybdenum grade recovery curves to forecast the circuit production for each of the first 10 years. INTRODUCTION Incorporating Geometallurgy into the design for a copper molybdenum separation is especially challenging due to the low mass of final concentrate typically produced during testwork. This is due to the low molybdenum head grades and high overall molybdenum upgrading expected. With molybdenum head grades of around 300 ppm, 1 tonne of copper circuit feed is expected to produce around 300 g of molybdenum concentrate in the plant. In the laboratory however, this may end up only resulting in a few grams. Obtaining sufficient mass for laboratory testing therefore requires starting with a large initial sample and careful testwork planning to maximise the utilisation of this mass. A testwork program was designed to investigate the effect of ore variability on the copper molybdenum separation circuit (here after referred to as the molybdenum circuit). Composite samples were selected to represent the various ore types, copper content, moly content, pyrite content and clay contents expected in the copper circuit feed. These samples were run through a bulk copper pilot plant to create the feed for the molybdenum testwork. The molybdenum testwork consisted of rougher kinetic flotation tests, rougher cleaner kinetic tests, rougher cleaner kinetic tests with regrind, stage flotation reactor (SFR) pilot plant tests, open circuit cleaner tests. These tests were used to calibrate the ore kinetics in the IGS-FLEET model for simulation of the molybdenum circuit performance.

Jan 1, 2019

By Charles G. Kemsley

IN view of the truly remarkable development of special steels for special purposes, the mining industry has for years looked forward to a time when steel makers would come forward with a product which would be an improvement over the straight carbon steel which has always held the field for drilling rock. In order to avoid making changes in conventional shop practice, the industry has perhaps been too insistent upon obtaining a water hardening steel which could be handled in the blacksmith shop by long established methods. As early as 1932 or 1933, an oil-hardening tool-steel of the one per cent carbon, plus chrome, plus molybdenum type was applied to hollow drill steel. It was known as a hard as well as a tough steel, and one that, theoretically, should give a good account of itself as a rock drill, if properly heat-treated; and so it has proved in actual trials. There were numerous deterrents which retarded adoption of this steel. For several years there has been a constant improvement in the drilling qualities of the straight carbon steel, and, considering the higher unit price ? of the alloy, there has been considerable doubt in the minds of some mining? men as to what advantage there might be in the use of alloy over straight carbon for hollow drill steel. The alloy being an oil-hardening steel, its adoption meant having two kinds of steel in the mine, for a time, and that spelled trouble, unless some positive means of identification was adopted. The hardening and forging ranges of alloy steel being narrow also made necessary much closer control over these operations if poor results were to be avoided. These thoughts undoubtedly prevented its use becoming wide-spread, for, as a matter of actual record, less than fifteen per cent of all the hollow drill steel used in Canada today is alloy.

Jan 1, 1942

By Gustavo Brunello

Motor Failure Rates and Cost ? Motor failure rate is estimated: 3-5% per year. ? In Mining, Pulp and Paper industry failure rate is up to 12%. ? Sometimes process downtime costs are higher than motor ?? Motor failures are due: Electrical 32% Mechanical 31% Environmental 36%

Feb 1, 2008

By Alireza Sarafi, Alijan Aftabi

"A hydrocyclic system used for mineral separation at the laboratory scale results in the separation of more than 90% of micas. The system functions only by water and might offer advantages where other methods of mica separation are not available. INTRODUCTIONVarious methods for the separation of micas, wollastonite, and garnet, e.g., screening, flotation, magnetic separation, and heavy liquids, are well known (e.g., Spence, 1929; Bates,I960; Gillson et al., 1960; Benjamin, 1971; Connor and Kolhnar, 1991; Collings and Andrews, 1991). However, where these methods are not available, the separation of micas by a hydrocyclic system developed by Sarafi and Aftabi (1991) may prove to be economic. This note is a brief description of the hydrocyclic system. Minerals and MethodFor testing the recovery of mica and other minerals from different rocks (garnet-mica schist, skarn, and pegmatite), standard minerals were tested first, and then rock samples. Pure mineral standards were prepared from coarse-grained samples of pegmatite, mica schist and skarn; the grain size of the minerals separated was more than I cm. The Hydrocyclic System and Mineral SeparationThe hydrocyclic system is shown schematically in Figure 1. In the first step in mineral separation, mineral and rock samples were crushed, milled, and then sieved to minus 250 plus 125 microns. Approximately 10 gm to 25 gm of the sample were inserted in decanter Dt • The motor pwnps, M1 and M2, were then turned on and the water valve rl was opened. At this time, water was flowing turbulently (30 em3/sec. to 50 em3/sec.) and minerals were separated according to their crystallographic nature, density and!or habit. The fInal separation was done by exit valves r2, r3 and r4, while the motor pwnps were still operating. After each experiment, the rt , r2' r3 and r4 valves were closed and the motor pumps were turned off. Following this, minerals in each decanter and the mter were collected, dried and weighed. The whole time of the experiment was optimized to be about 10 minutes, but varied from 3 to 15 minutes for micas and garnets. respectively."

Jan 1, 1994