Search Documents

By C Aldrich

In this paper, the use of topology preserving projections (Sammon maps) and neural networks are considered as a basis for the construction of multivariate statistical process control charts for a semi-autogenous grinding circuit on a gold plant. Comminution circuits are often operated in a quasi-steady state mode, where the mills tend to move rapidly from one state to another. This makes them difficult to monitor by conventional means designed for continuous steady state processes. As a first step, process data reflecting normal operating conditions are mapped with SammonÆs algorithm. Only one feature is extracted at a time and each mapping is captured by a neural network model to allow online application. Reverse neural network models are used to reconstruct the data prior to the extraction of subsequent features from the data residuals. The ensemble of neural network models can then be used for both fault detection and identification. The approach compared favourable with both linear and non-linear principal component models. It was able to detect two simulated fault conditions both with a reliability of 100 per cent, as opposed to 57.8 per cent and 95.2 per cent detected by an approach based on the use of linear principal component analysis and 84.1 per cent and 98.0 per cent detected by an approach based on non-linear principal components extracted with auto-associative neural networks.

Jan 1, 2007

By Ralph Sweetser

THE present state of iron blast-furnace practice is metallurgical rather than mechanical; the tend-ency is toward intensity rather than toward ex-tension. The engineers have built blast furnaces big enough and strong enough, and have furnished accessory equipment ample and accurate enough for the present state of the art, but the metallurgical side has been too much neglected; the metallurgist, rather than the engi-neer, is now the man on whom the present and future progress in iron blast-furnace practice depends. The mining engineer has produced great tonnages of iron ores, and has kept ahead of the demands for quantity; now the metallurgist is improving the chemical and physical qualities of this product by ore-dressing and other forms of beneficiation. For a generation or more, the principles of mechanics have had most influence in the advancement of iron blast-furnace practice; atten-tion now is given more to the principles of metallurgy, chemistry, and ore-dressing. This metallurgical progress is taking place inside and outside the blast furnace. The two great determining factors of the inside operation are the slag, and the rate of combustibility of the coke. Outside the fur-nace, the progress is, and will be, along the lines of more careful mixing of ores and iron-bearing materials, in the sizing and washing of limestone, and in the char-acter and structure of byproduct coke.

Jan 3, 1922

By C. E. Feltner

In a recent note, Meleka and Hiarris1 have reported that the application of a constant strain rate Ad a superimposed cyclic stress to 99.99 pct polycrystal-line zinc at room temperature produced some marked effects in the resultant behavior. In their work, one end of the specimen was pulled in tension at a constant strain rate while the other end was subjected to a sinusoidally varying load (20 cps) of constant amplitude. For the slower strain rates and the larger superimposed cyclic loads, it was observed that the specimen could withstand strains prior to failure more than 100 pct greater than the fracture strains obtainable in simple unidirectional straining. This ductility effect was interpreted as being due to a cyclic stress enhanced thermal recovery process. It is the purpose of the present note to describe some related effects in 99.99 pct polycrystalline cadmium. Specimens were loaded in repeated tension at room temperature with a stress of sufficient magnitude to cause failure in about 1000 cycles and the accumulated strain was measured as a function of cycles and time. In addition, some specimens were subjected to a steady creep stress equal in magnitude to the peak value of the cyclic tensile stress. The resultant strain-cycle and strain-time curves for both types of test are shown in Fig. 1. It should be noted that the static creep specimen failed in about one third the time required to produce a cyclic failure but in the cyclic test the specimen was able to withstand a uniform strain before necking more than 50 pct greater than that possible

Jan 1, 1963

By Jonathan M. Berkoe

Design engineers and metallurgists have long recognized the need for better methods to calculate a wide range of fluid dynamic effects in smelters and hydro- metallurgical facilities. Examples can range from the capture of fume in the secondary hoods of a copper converter, to controlling the thermal effects of heat transfer resulting from metal and slag flows within a furnace. With the advent of relatively low-cost, high- speed computers and user-friendly, commercially available computerized fluid dynamic (CFD) programs, these calculations can now be made. Use of CFD increased the understanding of both hot gas and molten flows in smelter facilities and has already led to design improvements in converter fume control hoods as well as improved management of heat transfer in electric furnaces. This paper will present two representative CFD applications to metals industry facilities. To date, Bechtel has implemented CFD results on four continents.

Jan 1, 1999

By H Waterer, L Zheng, T Kalinowski

The Hunter Valley Coal Chain (HVCC) consists of mining companies, rail operators, rail track owners and terminal operators, together forming the worldÆs largest coal exporting facility. In 2008, the throughput of HVCC was about 92 million tonnes, or more than ten per cent of the worldÆs total trade in coal for that year. The coal export operation generates around $15 billion in annual export income for Australia. As demand has increased significantly in recent years, and is expected to increase further in the future, efficient supply chain management is crucial. The Hunter Valley Coal Chain Coordinator Limited (HVCCC) was founded to enable integrated planning and coordination of the interests of all parties, thus improving the efficiency of the system as a whole. One of the many planning challenges faced by the HVCCC is that of annual maintenance planning. Different supply chain elements, such as trains, railway track, terminal equipment and load points, must undergo regular preventive and corrective maintenance, leading to significant reductions in system capacity (up to 15 per cent). However good alignment of the maintenance tasks can reduce their impact, and the HVCCC undertakes an annual process to ensure the impact of maintenance on the supply chain capacity is as small as possible. This is achieved in an iterative negotiation process between HVCCC and individual service providers. In the past, maintenance schedule optimisation was largely manual, which for the more than 1000 tasks involved is quite labour-intensive.In this paper the authors describe an approach developed at the University of Newcastle in partnership with the HVCCC to automate this schedule optimisation process. We will discuss our experience in applying exact (mixed integer programming) and heuristic techniques from mathematics and computer science to address the problem. This work is anticipated to lead to new decision-support tools for the HVCCCÆs capacity planning team.

Sep 26, 2011

By Kenneth E. Mathews, Ross D. Hammett, S. Byron Stewart

Theoretical models have been important throughout the history of engineering as a means of understanding past experiences and predicting future performance. Most modelling involves assumptions and a degree of idealization, and it is shown that it is important that the effects of these be recognized when making use of models. One of the principal concerns in underground mining is the effect of the excavations on the states of stress in the rock adjacent to these excavations. As mining proceeds to depth, stress concentrations may be sufficiently high to cause inelastic behaviour of the rock and yielding along joint surfaces. At this stage, attention must be given to the sequencing of the mining process to ensure that the ground control techniques applied are optimized. Increasing use is being made of computer modelling techniques to assist with design. Three case studies are presented to illustrate the application of computer models in mine planning and ground control. In all cases, the data available for design was limited, but the mines were fully developed and openings could be inspected. The mining methods studied are cut and fill stoping, room and pillar stoping, and longhole stoping.

Jan 1, 1983

This study presents a reservoir modeling approach to investigate the relative effects of different coalbed parameters on the migration of methane into development entries. A base coalbed reservoir model of a three-entry development section, where grids were dynamically controlled to simulate the advance of mining at a constant section advance rate, was created and calibrated for a Pittsburgh Coalbed mine in the Southwestern Pennsylvania section of the Northern Appalachian Basin. The values of coalbed parameters were varied to evaluate their effects on predicted methane emissions for various development distances. The results of these parametric simulations were then used to derive linear expressions relating these parameters to methane emissions into the workings. These models were analyzed to assess their significance and adequacy for predictive purposes. This work shows that coupling reservoir simulations with linear modeling yield a technique that can be applicable to different coalbeds. The reservoir parameters used by the linear models (coalbed thickness, pressure, sorption time constant, Langmuir parameters, permeability) can be determined by running relatively simple laboratory tests, such as adsorption equilibrium and permeability determination, on coal samples obtained either from the mining operation or from the exploratory boreholes drilled ahead of mining.

Jan 1, 2008

By V. Roze, M. E. A. Robertson, C. J. Sharland

Current Techniques - Fire Assay Fire assay of gold has been used for several thousand years and is a well known technique to the traditional assayer. Very few assayers wish to use the method but to date there has been no viable alternative. Fire assay is expensive in labour, time, capital and running costs. Workers can be exposed to hazards and uncomfortable working conditions such as lead fumes, heat stress, dust and eye strain. The high capital cost of putting up the building and equipping it, together with running costs which increase annually at a prohibitive rate makes the technique less attractive each year. The main requirement in most assay laboratories is to maximise throughput economically. This is difficult with fire assay as it is impossible to automate. Each sample for analysis is handled by at least 12 people and each person represents a potential source of error. Record keeping and report generation has to be done manually which is both slow and prone to errors. Replacement of staff requires a considerable training period and with a large turnover of personnel this presents yet another problem. The presence of silver and dross in the gold sample affects the accuracy at low gold values. X-ray fluorescent analysis Two methods of XRF instrumentation have been used with limited success. The first type is a laboratory instrument employing L X-rays. Small samples of a few grams, which are not representative,

Jan 1, 1988

By Kenneth J. Reid

The application of computers for control purposes in hydrometallurgical processes is not as well-established as in mineral processing plants. This is not only due to the comparatively new development of large-scale application of hydrometallurgy and the relatively small number of operating plants, but also due to some of the basic dynamic characteristics of hydro- metallurgical processes., Basically there is little difference in the development and application of digital control techniques in hydrometallurgical plants from those techniques widely in use throughout the chemical process industry and in nonslurry applications; particularly for those cases involving liquid streams only, the primary sensing elements are essentially the same. On-stream X-ray fluorescence analysis is a well-established method for determination of metal content in the process streams, and standard equipment is used for the measurement and control of pH, flow, density, temperature, pressure, and other variables. Nevertheless, the high proportion of corrosive slurry streams at elevated temperature and pressure poses a challenge to the instrument engineer concerned with developing long-life, reliable instrumentation. One major difference and potential problem area in developing control systems for hydro- metallurgical processes is related to the very long time constants which are common in the larger plant. In these circumstances, it is important to define the overall control objectives very carefully and be prepared to allow for considerable drift in intermediate variables. In larger projects with many control loops the general advantage of lower costs compared to conventional analog instrumentation can be confidently expected. Also in most cases, the process analysis necessary to design and specify a comprehensive computer control system and the increased operator involvement in implementing digital control systems yield additional benefits due to the increased understanding of the dynamic characteristics of the process.

Jan 1, 1979

By P. P. Goudkoff

UNDER the name of the Russian Far East we understand the territory occupied by the Amur, Maritime, Sakhalin and Kamchatka Provinces, the total area of which is about 918,000 square miles. The mining industry took its first step in the Russian Far East in the second half of the 19th century, and until the end of this century was restricted to gold placer mining. In the last years of the 19th century, the extraction of coal and lignite was begun, and in 1910-11 the exploitation of silver-lead-zinc mines. The conditions of the Far Eastern mining industry during the last few years may be outlined by the following figures. Annual production of gold mines was about 54,000 lb., valued at $13,000,000. During the period of 1910-15 a considerable increase of placer-gold mining by mechanical processes took place; in 1910, by means of dredging and excavators, 500 lb.. of gold were obtained; in 1914, 2740 pounds. Annual output of coal; lignite and. silver-lead-zinc ores was: 1911 1920 Bituminous and semi-anthracite coal, long tons 270,000 316,000 Lignite, long tons 66,000 205,000 Silver-lead-zinc ores, long tons 28,400 94,500 In other branches of mining in the Far East, only the quarrying of building stone and clay may be men-tioned, the total yearly output amounting to about $500,000 in value. The principal gold mines are in the Amur Province, in the systems of the Zeia and Bureia rivers, and in Kamchatka Province, near the town of Okhotsk. The above-mentioned tonnage of coal and lignite was extracted from the mines in the Maritime Province, near Vladivostok. The only silver-lead-zinc mine worked was that of Tetiukhe, near the Tetiukhe Bay, 240 miles northeast of Vladivostok. The ores of this mine were exported to Germany, Belgium, and Japan, partly raw, partly calcined.

Jan 6, 1922

By Lance A. Roberts

The Federal Highway Administration (FHWA) has mandated the use of the AASHTO Load and Resistance Factor Design (LRFD) specifications for all new federally-funded transportation infrastructure projects beginning in October 2007. This includes the design of all geotechnical-related elements, such as deep foundations, which have historically been excluded from the LRFD requirements. The basis of the LRFD specifications is the application of a probabilistically computed parameter called a resistance factor to the foundation capacity in lieu of the traditional, and often arbitrary, factor of safety. Therefore, the process of computing resistance factors for use in the design of deep foundations is extremely critical to ensure a safe, efficient, and consistent design methodology. In this paper, various methods to calibrate resistance factors for use in the design of deep foundations are examined. The methods that are currently utilized to calibrate the resistance factors reported in the AASHTO LRFD specifications may be classified as the allowable stress design (ASD) method and the reliability theory calibration method using static capacity prediction techniques. These methods are compared to a ?t-z? model design methodology. The current methods focus exclusively on the strength limit state design of deep foundations. The advantage of the ?t-z? model methodology is that it can be utilized for both strength and service limit state design. In addition, the results show that resistance factors vary based on geological setting, construction method, failure criteria and deep foundation type. This variability in the resistance factors should therefore be integrated in the design to ensure that the objectives of the LRFD approach are satisfied.

Jan 1, 2008

By Monte J. Pool, Rudolph Speiser, George R. St. Pierre, William T. Ebihara

Standard free energies of formation of WO,, W O W20058 and WO3from oxygen and the lower oxide or tungsten have been determined in the tempel-ature range of 700° to 1220°C. A tentative W-O phase diagmm is presented based on the evidence that the oxide phases, "W18O49" and "W20058", are unstable at l~atm total pressure, below 585° and 485°C:, respectively. A HE need for a complete quantitative treatment of the phase equilibria of the W-O system has been establishedby metallurgists during the past few years. In an attempt to improve the endurance of tungsten exposed to oxidizing atmospheres at high temperatures, many research groups have studied the kinetics of tungsten oxidation. The reverse process, the reduction of the tungsten oxides to metal, has also been investigated to some extent. Regardless of the particular interest, metal to oxide or oxide to metal, a knowledge of the phase equilibria of the W-O system is required to properly interpret all of the kinetic observations. It should be possible to state the stable phases of the W-O system at any given values of the chemical potential of oxygen and the temperature.

Jan 1, 1962

By Chester W. Washburne

At present it is impossible to estimate closely the amount of oil obtainable from a given area of land. However, after the completion of a few properly distributed prospect wells, one can calculate the approximate yield of the sands penetrated, with a probable error of say 50 per cent. Even rough predictions of this kind are of value in large operations. Without wells upon it, one can never be absolutely certain that an area will produce any oil whatever, and one can only indulge in wild guesses of the probable productive area, before the geological structure has been carefully contoured and the top of the basal water plotted on a map. This requires wells. At times it is desirable to form an idea of the possible yield of oil in advance of drilling, as in cases where the price asked for undeveloped land seems too high, but where the probability of finding oil is furnished by neighboring development, supported by the geology. In such cases the most one can do is to make an estimate of possible maximum yield, which may of course far exceed results, and which is of use mainly in the prevention of excessive initial investment. The method is very significant in a general way, as in regions like the Bighorn Basin, Wyoming, where the productive sands in the Cretaceous shale are so few and so thin and fine that they cannot develop very great production, although the high quality of the oil probably will counterbalance this defect sufficiently to give local profit in favorable areas. Reservoir Capacity of Sands.—The porosity of a stratum is the measure of its maximum reservoir capacity for liquids and gases. The porosity may be determined experimentally and must be used in connection with the total volume of the sand.

Jan 1, 1915

By R. W. Raymond

On the 27th day of March last, the day on which the Florida sessions of the last meeting of the Institute began in Ocala, occurred the death of one of its most distinguished honorary members, Bergrath Franz Posepny, of Vienna, formerly professor of the science of ore-deposits in the mining school at Przibram, Bohemia. Prof. Posepny had greatly increased his fame among American mining engineers, besides laying the Institute under special obligations of gratitude, by the elaborate, brilliant and suggestive treatise on " The Genesis of OreDeposits," which he contributed to the International meeting, held at Chicago, in 1893. This essay, constituting the first complete publication of the substance of his course of lectures at Przibram, enriched and perfected by the matured results of his investigation and reflection down to the date of its cornpletion, was a free gift of almost unprecedentecl value to the society which had distinguished him by honorary membership. In estimating the generosity of the author, it must be borne in mind that the copyright of such a work, the fruit of years of study and of practice as an instructor, is of no little value to a European professor, and constitutes one of the legitimate rewards of his (otherwise not highly-paid) labor. Moreover, Prof. Posepny performed under peculiar difficulties

Jan 1, 1896

In underground ore mining of copper ore, lead-zinc ore, tin ore and fluorspare in German Democratic Republic geological situation is characterized by ore deposits of small thickness and great variability of ore grades and mining conditions. For long range mining strategy and short-term planning of production a model was developed fitting to make optimal economic mining decisions. Starting point of the model is a cost function representing all direct costs per unit recoverable metal in concentrate for each stoping unit (ore reserve block). In the costs function the results of regression analysis of ore grade and metal recovery in concentration plant were considered. All stoping units (ore reserve blocks) of a mine were arranged to growing direct costs. The relationships of recoverable metal, ore tonnages and corresponding costs were plotted in a diagram. The data processed by- computer and the graphic representation of the results make possible 1. economic evaluation of stopes and ore reserve blocks, 2. estimation of cut off grade, 3. optimization of works planning.

Jan 1, 1977

The impulsive loading tests under confining pressure were performed to investigate the dynamic fracture properties of rocks in a triaxial stress state. The apparatus in this study consists of the Split Hopkinson pressure bar, a pressure vessel and hydraulic actuator. Magnetite-quartzite and limestone were selected for the test. The stress-strain curves and fracture strength under confining pressure obtaind in the tests were compared with those of the conventional triaxial tests. Finally it is concluded that on the effect of dynamic that on the effect of dynamic compressive strength of rock, dynamic elastic coefficient of rock and confining pressure etc. on the blasting result

Jan 1, 1990

By H. W. Schreiber, H. R. Babitzke, R. E. Rodriquez, Patterson J. A., C. D. Masters, J. J. Schanz, R. B. McCammon, E. A. Noble, P. M. Krishna

A five-year effort by the ASTM subcommittee E10.14 on uranium resource evaluation has culminated in the approval of E901-82, Classification System for Uranium Resources (ASTM Standard E901-82, April 1983). The standard classification was prepared with participation of representatives from the US Department of Energy, the US Department of Interior's Geological Survey and Bureau Mines, utilities, and consultants. The classification a resource is based on level of knowledge and the economic producibility of the resource at a given time. Alphanumeric identifiers are used rather than words to avoid confusion with existing terms, definitions, and classifications. An example demonstrates the dynamic nature of the process of classification and attests to the necessity of addressing both the aggregate needs of broad-scale resource planning and the specific needs of individual property evaluation. Problems that remain infixing the classification a given uranium resource include uncertainty in estimating the quantity of undiscovered resources and resolving differences that may exist in deciding when the drill-hole spacing is adequate to determine the tonnage and grade of discovered resources.

Jan 1, 1985

By M. Lahtinen, H. Takala, B. Saxén

Oxidic zinc fume dusts are formed in electric arc furnace (EAF), Waelz furnace operations, and also in the Top Submerged Lance (TSL) furnace of Outotec Ausmelt. These dusts are normally considered as secondary raw materials for zinc metal production, as they are mainly originated from steel scrap recycling or zinc plant residue processing. The latest development in zinc fuming is direct smelting of zinc concentrates in the Ausmelt TSL furnace and hydrometallurgical treatment of produced fume dust. This technology is for primary zinc production like roasting and concentrate direct leaching, as concentrates are directly treated. The various oxidic zinc raw materials differ in composition and content of impurities. Thus, they require different hydrometallurgical processes to produce pure zinc metal. In addition to zinc these dusts typically contain also lead, silver and indium, which can be recovered and can increase recovered metal value significantly. From the impurity point of view the most relevant elements in the fume dusts are iron, arsenic, fluoride, chloride and cadmium. This paper describes hydrometallurgical processes to treat Ausmelt TSL fume dusts to extract valuable metals at high recovery and also to control and separate various impurities efficiently.

Jan 1, 2014

By Michael G. Nelson

Underground coal mining practice often requires that some coal be left in place at the seam boundaries after mining, for purposes of ground control, quality control, machine guidance, or good operating practice. Many methods for measuring the thickness of boundary coal have been developed, but only the natural % R gamma background method as bee used in routine operations. Use of this method is constrained by the level and consistency of the natural gamma background in the bordering strata, the accuracy of the gamma detection instrument, and the elevation profile of the seam boundary being detected. Computer simulations for three commercially available gamma detection instruments, designed for mine use, showed that measurement of boundary coal thickness by the natural gamma method can provide satisfactory input for horizon control systems when the gamma detector is large, when the sensor is mounted on the mining machine, and when the elevation profile of the seam boundary does not change rapidly in the direction of advance.

Jan 1, 1993

By A. Filzwieser

The highest current density suitable for the electrowinning process is between 50 to 60 % of its limiting current density. One of the most important influences on the limiting current density is the thickness of the hydrodynamic boundary layer near the electrode surface. Therefore, the fluid flow in a copper electrowinning cell is calculated. The numerical simulation - using the CFD-software package FIRE® - considers all three different types of the fluid flow, namely, the natural convection, the forced convection by electrolyte circulation and the forced convection by electrochemically induced gas stirring. The simulation of natural convection is based on the different density values in the boundary layer at the electrode surface given by a density/concentration correlation. The copper concentration - linked with the current density by Faraday law - is solved by an additional transport equation. The simulation of the forced convection by the electrochemically-induced gas stirring is done by using a real two-phase calculation. That means that all differential equations are solved again for the second phase. The numerical solution of the fluid flow field is compared with the results of LDA- measurements, which were done in a special cell. The software for the LDA enabled a calculation of the void fraction possible and therefore a rough estimate of the mass transfer coefficient through the prevailing gas bubble induced convection can be made by assuming the influence of relative bubble volume.

Jan 1, 1999