Search Documents

By G Anderson, V Lawson, P Voigt

Glencore operates zinc concentrators at both Mount Isa and McArthur River. Both orebodies have a history of complexity and required the development of new technologies to maximise the value from these long-life assets. The Mount Isa orebodies were plagued since operations started in 1931 with complex mineral textures making separation challenging and many areas uneconomic. Examples are given in Figure 1 with different grain sizes from the same orebody in different areas. The McArthur River deposit, located in the Northern Territory, was discovered by Mount Isa Mines geologists in 1955 (Logan, Murray and Williams 1990). The extremely fine-grained nature of the ore was quickly recognised and the difficult separation of the minerals was soon obvious. The geographical, metallurgical and economic factors resulted in the orebody not being developed until the technologies for the mineral separation had been developed. The development of the IsaMill allowed for an economic mineral separation flowsheet to be developed and ore treatment commenced in May 1995 (Nihill, Stewart and Bowen 1998). CITATION: Lawson, V, Anderson, G and Voigt, P, 2018. Developing technologies for a complex orebody, in Proceedings Complex Orebodies Conference 2018, pp 116–118 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 21, 2018

By L. C. Brichta

The Melrose zinc-lead. deposit is in an area of structural disturbance of sedimentary rocks in the Melrose mining district in Cherokee County, Kans., 25 miles west of Joplin, Mo., adjacent to the northwest limits of the Tri-State district. Mining and milling in the Melrose district has been sporadic and. very limited, largely due to the high cost of exploration through great thicknesses of barren rock, 350 to 440 feet overlying the ore deposits. No mining has ever been done on the Melrose zinc-lead tract, but it has been drilled extensively. The drill-hole logy indicate extensive structural deformation, silicification, and zinc-lead deposits in the Boone limestone-chert formations of the Mississippian series, which is overlain by thick-bedded Pennsylvanian shale and sandstone. Analyses of samples from Bureau of Mines drilling indicate the ore minerals are sphalerite and galena with no evident oxidation. During the period from November 2, 1949, to March 13, 1950, the Bureau of Mines investigated the deposit by churn drilling and sampling 23 holes aggregating 10,405 feet of bore. Factual data on the investigation and logs of holes drilled by private parties and the Bureau of Mines are presented in this report.

Jan 1, 1951

By M. Piard

Fires and explosions are not likely to originate in clean gobs above pillar faces steadily driven through a solid panel in order to effect a complete extraction. As. to the storage which such a process implies where the market is intermittent, its cost will never equal that of abandoning large pillars in a fully developed property. Much less will that cost, thanks to the big sizes obtained, ever approach that of cleaning wet and dirty crushings robbed at great expense from small pillars exposed to many years of roof pressure. Even thick seams under bad roof may be fully extracted without the help of filling material. Fatal accidents from falls of roof owing to the great height of excavations, losses in coal, and the danger from fire, have compelled operators in South Staffordshire, England, to discontinue the stall-and-pillar system at one time supreme in the district, and to successfully adopt the following plan: A retreating' longwall is run along the top coal, as the first working; the bottom coal is taken in the same manner, after the fallen roof has become sufficiently consolidated to allow the mine to be reopened. A similar plan is suggested in this paper in order to deal with the outcropping and pitching seams found in the Kootenay measures.

Jan 1, 1930

By Bruce H. Mason

The exploration for a commercially acceptable deposit of limestone is consid¬ered by many to be a relatively easy task; however, trying to locate a zone of high calcium limestone of economic thickness within a formation, which outcrops only in a narrow band varying in width from a quarter mile to ten miles, adds considerably to the normal problems experienced during an exploration program. The existence of varying glacial and karst features in the outcrop area intensified the problems encountered and complicated the interpretation of the economic value of the deposit. This paper highlights a recently completed five year exploration program during which approximately 3,900 acres of ground in seven civil townships were evaluated as potential sources of high-calcium limestone. Details are given with respect to the final site chosen and the methods used in determining the suitability of the site. This site exhibited many karat features and several methods were utilized to determine the effect of these features on the gross characteristics of the deposit. Aerial photography for both mapping and interpretation were utilized. Black and white panchromatic photography, black and white infrared and color infrared photography were flown. Resistivity, diamond core drilling and close-interval angering were used to supplement the information obtained from the aerial photography.

Jan 1, 1974

By E. A. Noble

In the Western Coal Region, successful reclamation of surface-mined areas includes the protection of groundwater sources. To accomplish this, a complete understanding of the geohydrology of the affected area is imperative. Successful mine design also calls for careful attention to geohydrology. Because the Western Coal Region is semi-arid land used mainly for farming and grazing, groundwater is a particularly sensitive topic. It is especially sensitive where the lignite scheduled for mining is itself the aquifer supplying the farms and ranches. Such is the case throughout much of the Western Coal Region, particularly where there are no near-surface sand and gravel aquifers that are commonly found in the glaciated portions of the Region. Just as mining can adversely affect groundwater supply, so can groundwater adversely affect mining. Mine flooding and highwall instability are examples of this. But, mine operators must realize that procedures they use to remedy geohydrologically-related mining problems can have a dramatic effect upon ground- water conditions in areas adjacent to the mine. A logical precaution is to require the acquisition of complete baseline data prior to making the first cut. We recommend that the mining plan not only incorporate the reclamation plan, but that they be developed as one plan.

Jan 1, 1976

By Michael Taylor, Thomas W. BoBo

Currently two of the most important characteristics of construction aggregates are (a) the particle size distribution (psd) and (b) the “shape” of the aggregate particles. The long established test methods and standards for these properties require substantial resources and most suppliers can take samples and conduct tests at intervals measured in days (psd) or even months (shape). Furthermore these tests require time: so the results are available only after material has been stockpiled or shipped. In the rare cases where problems are found the out-of-spec material has often been shipped to points unknown. There are substantial advantages to be obtained if the psd and other geometrical properties could be obtained from the material as it moves through the production processes and if the results could be reported back within seconds or minutes to Plant Control Managers. This paper describes a single on-line system used to evaluate production of crushed material in a production quarry. The system consists of an enclosure (containing the camera and lighting etc.) mounted over a production belt. Images are fed by fiber optic cable to a nearby control room where the data are analyzed; stored; displayed; and made available for integration into a process control system. Numerous forms of display can be created according to the needs of the producer. Once programmed by an automation engineer all functions of operation are fully automated. The system has permitted over 400 tests per hour with a feed back time less than 2 seconds. Sufficient geometrical measurements are made so that many different “shape” factors can be computed on a similar time scale. The paper summarizes the results from approximately 100 shifts of about 8 hours each and compares the imaging method with the current standard hand methods. Study of the data generated showed that, when combined with additional plant information, the system can be used to generate additional useful control information.

Jan 1, 2005

By J. R. Linney

APPROXIMATELY 5,788,000 long tons of crude ore was produced by the Eastern magnetite industry in 1946, or a drop of 26 per cent compared -with 1945. Decrease in production -throughout the industry varied from :8 per cent for the mine with the lowest reduction to 49 per cent for the mine -with the highest reduction. At the Chateaugay mine, operated by :the Republic Steel Corp. at Lyon Mountain, N. Y., 426,900 long tons of crude ,ore was produced, yielding 129,500 long tons of concentrates or a reduction of 46 per cent compared with the previous year. Owing to a strike the mine was closed down from Jan. 21 to July 29. At the Mineville mines, also operated by Republic Steel, output was 755,000 long tons of crude ore or a reduction of 49 per cent compared with 1945. The yield from the crude ore production consisted of 51,500 long tons of lump ore and 380,000 long tons of concentrates. The Mineville plant did not operate from March 17 through July 28, on account of the strike. The Fisher Hill mine, which produced 93,000 long tons of crude ore, included in the above figures, has not operated since March 17, 1946.

Jan 1, 1947

By T. Holmes, J. Middlemiss

"Agnico-Eagle (AE) has been operating their Kittila Mine in Northern Finland since 2008. The mine has experienced considerable ore flow problems with ratholing, bridging and freezing in their mill feed bin since start up. In the first winter of operation, the entire central portion of the contents of the cylindrical flat-bottomed 2000 tonne mill feed bin froze. Flow was erratic and could only be achieved through small diameter ratholes. To maintain a reasonable, albeit limited, feed to the mill, the bin had to be kept less than ½ full and a portable steam generator was brought in to melt the frozen agglomerates that frequently blocked the outlets. The combination of limited surge capacity and frequent outages required the crushing plant to operate 24/7 and resulted in inconsistent production levels and high operating costs.AE had planned to expand the processing plant to increase throughput by about 30% when they transitioned from open pit to underground ore. These plans presented an opportunity to improve the material handling.To avoid the problems experienced with the original mill feed bin, AE enlisted Jenike & Johanson Ltd. (J&J) to test samples of the ore, and based on the measured flow properties to provide design recommendations to ensure that the new mill feed bin would operate reliably. The new 2000 tonne expanded flow mill feed bin and apron feeder were installed in 2014 and are now supplying almost 93% of the total mill feed with the remaining coming from the original bin.IntroductionIn early 2009, AE engaged J&J to visit their then recently opened Kittila mine and mill in Northern Finland (Figure 1). Over the first half-year of operation, production had been restricted in the ore handling and storage system by flow problems such as ratholing, bridging and freezing including blockages of chutes and hoppers as a result of frozen fines buildup and large frozen agglomerates. AE wanted to increase production in the existing open pit mine and mill feed operation and also wanted to avoid similar ore handling problems when they commissioned the new underground mining system under development at that time."

Jan 1, 2017

By S. D. Thompson

Since 1965 several researchers have independently developed computer models for designing gathering belt and surge bin capacities in underground coal mine belt systems. Table 1 summarizes the models applicable to U.S. mining systems. Most of the models, employ the Monte Carlo simulation technique. Basically, the belt loading at each loading point is randomly produced from probability distributions. In some cases, a combination of deterministic calculations and probability distributions are used to simulate the loading. Once the loading is simulated, it is systematically moved through the belt system in a deterministic fashion based on belt speeds, belt lengths, and storage bin discharge rates. The coal flowing from each loading point is tracked as it moves through the belt system. When the loading from tm different sources combines on a belt or in a surge bin, checks are made to determine if capacities have been exceeded. If capacities have been exceeded the amount of spillage is calculated. At the end of the simulated time period, the cumulative spill- age is used to evaluate the belt and surge bin capacities. Naturally, a considerable amount of spillage indicates the capacities are unacceptable. Acceptable capacities are generally indicated by minimal spillage. Numerous trial simulations are often necessary to find the acceptable capacities.

Jan 1, 1987

By B. Lueger, D. Shen, M. Sedighy, R. Rand, R. Saleh, J. Cheung, M. Rifai, I. Khan, W. Sauri, D. Setianto

The world’s first active power demand stabilization system for a 3-electrode electric arc smelting furnace was commissioned at PT Vale Indonesia in December 2012 on Furnace No. 2 using an SPLC (Smart Predictive Line Controller). The SPLC technology in combination with the Furnace No. 2 rebuild and other upgrades has enabled PT Vale to increase the electrical input to the furnace by up to 20% on a stable and sustainable basis by eliminating peak powers that overload and trip out the electrical equipment, and by reducing power unbalances. The SPLC technology maintains smooth arc power using a continuously variable thyristor controlled series reactor. The SPLC compensates variations in the furnace arc with predictive changes in the circuit reactance at a speed of up to 60 times a second. The system started up quickly and has been shown to be easy to operate and maintain by PT Vale. PT Vale’s Operations Team has been testing the new system at set points of up to 90 MW on the newly re-built Furnace No. 2. Over the last 12 months, the throughput of Line 2 has increased by approximately 10% on an annual average; however, debottlenecking in other design areas of the operation is required in order to achieve 90MW operation on a sustainable basis. This outstanding success results from a combination of the furnace and other upgrades, the SPLC active power stabilization, and the skill of PT Vale in quickly adopting new technologies and maximizing the combined benefits of each technology in their operation.

Jan 1, 2014

The October General Meeting of the Institute was held in Kelvin House, Johannesburg, on 15th October, 1969, at 4.30 p.m. Mr J. K. E. Douglas (President) was in the Chair. There were also present 17 Members, Messrs: M. Barcza, R. C. J. Goode, J. A. Nixon and V. C. Robinson (Council Members), M. L. Fitzgerald, D. F. H. Graves, C. G. Hinds, S. D. Hill, J. S. Hall, D. F. Malan, T. Meyer, W. pyne-Mercier, K. S. Rae, S. G. Taussig, L. W. P. van den Bosch, Dr A. Whillier, C. H. Wyndham. Four Associate Members: Messrs O. Davel, M. J. Martinson, M. P. J. Sandys, T. G. Harington. One Student: R. P. W. Henrard. Three Visitors: Messrs A. V. Pike, B. Schmitz, A. P. Raats. Secretaries: D. C. Visser and S. de. Gersigny. Total present: Twenty-eight. OBITUARY The President: "It is my said duty to announce the death of Pieter Ulrich Fischer, Member, who joined the Institute in 1961 and passed away on 24th June, 1969. As a mark of respect to the memory of the deceased and in sympathy with the bereaved I would ask you all to rise and observe a few moments' silence." MINUTES The President: "The minutes of the last meeting will only appear in next month's Journal and I would request that the confirmation of these minutes be held over." MEMBERSHIP The President: "I have much pleasure in announcing that the names of the undermentioned candidates, having been published in accordance with By-Law 5.2.2, Council has elected them to membership of the Institute in the following grades:" Members: Alexander Edward O'Meara, Anton Leopold Melvill, WilIiam Harold Wise, Eric John Dunstan, Frank Nel, Lubertus Hendrikus Jakob Poortier. Associate Members: Michael Irvine Britten, Selwyn Wilson Lange. James McCormack, Christiaan Lodewyk de Jongh, WilIiam Andrew Hustrulid, Reinhold Johannes Ortlepp, Felix Henry Lancaster. Graduates: Willem Adriaan Gericke, Hamish David Sneddon Miller, David lan Ossin, Robert Antony Layboume, Ewart John Shillaber, Colin Hewitt Wiggett. Affiliates: Bernard Andre Thomas, Waiter Arthur Wood, Andrew John George Paschalides, David John Perry, Alex Richardson, Gert Blekker Kotze. Student: Matthys Jacobus de Beer. MEMBERS TRANSFERRED TO A HIGHER GRADE From Associate Member to Member: David Alphonso Viljoen. From Student to Associate Member: Michael Charles Carruthers. "I welcome the newly elected members to the Institute and congratulate the members who have been transferred to a higher grade." PAPERS FOR PRESENTATION The President: "Professor Wyndham, may I call on you to deliver the paper entitled 'Factors affecting the mechanical efficiency of men shovelling rock in stopes', written jointly with J. F. Morrison, J. H. Viljoen, N. B. Strydom and A. Heyns." After the presentation of the paper the President said: "Gentlemen, I would like to thank Dr Wyndham and his colleagues for once again reporting through the medium of this Institute the results of the very important work they are carrying out at the Human Science Laboratories. It is most appropriate that this work which is leading to a better understanding of the capabilities and efficiency of the human body under the arduous conditions encountered underground should be reported to this forum. In this way our members, who are actively engaged in the job of mining, can firstly learn something of this research and secondly, can, with their practical knowledge of conditions underground, contribute to what has been learnt and possibly pose new problems for our scientists in the laboratories. The scientist should always have the opportunity of subjecting his research to the scrutiny of people engaged in practical operations. By the very nature of our gold mines we are unfortunately heavily dependent on a large labour force and while mechanization has been introduced wherever feasible, we are likely to continue to be heavily dependent on this labour. In accepting this fact we can be proud in this country of what has been achieved, not only in obtaining the best efficiency from labour but in ensuring that their working conditions are reasonable and that their health and well-being are adequately cared for. When one visits our gold mines, one cannot help being impressed with what is being done to ensure that we have a satisfied, happy and healthy labour force. The Human Sciences Laboratory with its scientific approach has assisted greatly in promoting a better understanding of the problems associated with labour under the conditions experienced on our mines. Thank you." Mr M. Martinson then contributed to the discussion. (This contribution will appear in a later issue of the Journal.) "Professor Wyndham, may I call on you to deliver the paper entitled 'Acclimatizing men to heat in climatic rooms on mines,' written jointly with Prof Strydom." After Dr Wyndham had presented the paper the President said: "Thank you Dr Wyndham for so ably describing the work associated with the acclimatization procedures on our gold mines. It has certainly been an eye-opener to those of us not familiar with this work and I believe it is very important that everyone and not only people in the gold mines should know about it. It is fortunate that all industries do not have the same arduous heat conditions experienced on our mines but other industries such as our chemical and metallurgical industries do have other problems relating to the adjustment of labour to the work situation. The scientific approach employed by the gold mines in tackling this problem is an example to these industries of what can be achieved by logical

By V. A. Leach, Lokan. K. H., S. B. Solomon, R. S. O’Brien, L. J. Martin, K. N. Wise

INTRODUCTION The development during 1979 of a relatively small, but high grade (10,000 tonnes uranium at an average grade of 2 per cent), uranium ore body at Nabarlek in the Northern Territory, Australia offered an excellent opportunity to obtain detailed radiation data for an open cut mine operating during the dry season. The ore body (Queensland Mines Limited-1979), which was completely extracted in a period of four and a half months, consisted of a vein type deposit dipping at 30 to 45 degrees and contained a central core of pitchblende in massive and irregular pods, surrounded by lower grade fine grained disseminated pitchblende. Mineralisation extended from the surface to a depth of 72 metres over a length of 230 metres with an average but variable thickness of 1D metres. Ore near the surface had been heavily weathered and complex secondary minerals were formed which had dispersed from the main vein. Mining was carried out with large earth moving equipment. Overburden and weathered surface ore were removed initially with scrapers. At greater depths bulldozers were used to rip and assemble ore and rock at each level, and these were removed by large trucks to the ore and waste rock stockpiles. Where necessary, blasting took place during shift changes each evening. Mining was essentially continuous with two ten hour alternating shifts working for thirteen days out of fourteen. At the completion of mining a relatively small excavation (335m x 185m x 70m) remained, and this will serve as a tailings repository during the milling phase. FIELD MEASUREMENTS The inhalation of radon daughters, arising from the radioactive decay of radon gas is well established (Archer et. al. 1973) as a potential hazard in the uranium mining industry. Control over radon and its daughters to ensure that recommended exposure limits are not exceeded is achieved by providing adequate ventilation, and under normal circumstances natural ventilation from an open pit should be sufficient. However, during the dry season it is not uncommon for stable atmospheric conditions, with little horizontal air movement, to develop - particularly at night - and significant radon daughter concentrations may accumulate. Throughout the entire mining period measurements were therefore made of radon and radon daughter levels at representative locations within the pit and on the ore stockpile as it developed. Initially these measurements were carried out manually, using the Rolle method for radon daughters, (Rolle 1972) and .scintillation cells (Lucas 1964) or a two filter tube for the determination of radon (Thomas 1970). For the latter half of the period however, a continuous recording instrument, developed within the Laboratory was used to provide a detailed record of radon daughter levels within the pit. At the same time, continuous readings of wind speed and direction, and vertical temperature gradient between 10 and 3D metres were recorded on a 30 metre meteorological tower, situated 800 metres from the pit. Radon Emanation Rates It is evident that radon and radon daughter concentrations depend on the grade, or more particularly, on the surface radon emanation rate of the ore which is exposed. Accordingly, as the mine progressed, detailed measurements were made of both of these quantities. The surface emanation rate of radon was determined for each ore bench as it was exposed by placing an extended array of canisters, filled with freshly degassed activated charcoal, face down on the ore for a known time. These canisters, which had previously been calibrated in the Laboratory, adsorb radon with high efficiency, and the total radon adsorbed is measured after retrieval by detecting the gamma rays from the trapped radon daughters (Countess 1977). At the same time, as each canister was placed in position, a measurement of the local ore grade was made for each location. This was achieved with a calibrated sodium iodide scintillation detector, adjusted to detect the 609 keV gamma ray from the isotope 2148i, a decay product of radium. Finally, measurements were made of the radiation field 1 metre above the surface, with a gamma ray survey meter, which was calibrated in the Laboratory. The relationship between the scintillator count rate and ore grade was determined by comparing the scintillator output with the gamma monitor, and relating the latter measurements to ore grade (Thomson and Wilson 1980). It was observed that while emanation rates and ore grades varied widely, the ratio of emanation rate to ore grade was in general fairly stable. A plot of this ratio is presented as a function of depth below the original surface in Figure 1. For most observations the ratio is constant at a value of 80 Bq m-2 s -1 per unit ore grade, where ore grade is expressed as percentage of U308. At the surface however, where the ore was weathered, the ratio was about a factor of three higher, and at two particular depths, where high grade pitchblende was being removed, it was very much lower. This was not unexpected as earlier Laboratory studies of drill core samples from Nabarlek had indicated that the emanation coefficient (the fraction of radon produced within the ore which escapes from the mineral particles) decreases with increasing ore grade.

Jan 1, 1981

By Aurel Goodwin

MSHA is the regulatory Agency which administers the Federal Mine Safety and Health Act. Although MSHA's principles largely derive from this Act, they do not exculsively derive from it. One might expect that our principles are contained in the regulations we have published to implement the Act. To a degree, this is true; but the regulations we have now were developed and published before the current Act became effective. For this reason, as well as others, there are some principles that we support which are not evident in our regulations, or in the Act. One of these is the ALARA principle. This principle requires that all unnecessary exposure be avoided; that is, all exposures should be kept "As Low As Reasonably Achievable." Almost all professionals in radiation control can agree with this broad principle; but it would be difficult to get agreement on a regulation which would translate the principle into industry practice. Rather, our regulations specify a limit to exposure for any individual. For radon daughters the current limit is four working level months per year. By having an explicit limit such as this, the ALARA principle often becomes lost and the limit becomes the goal. The principles contained in the Act are equally broad in scope. The principles apply not only for radiation protection, but also for toxic substances and harmful physical agents. The Act states that standards for such substances or agents shall most adequately assure on the basis of the best available evidence that no miner will suffer material impairment of health or functional capacity, even if such miner has regular exposure to the hazards dealt with for the period of his working life. The Act also provides that, although protection of health must be our foremost concern, other considerations shall be the latest scientific data in the field and the feasibility of the standards. Our regulations reflected similar principles when they were developed and promulgated. From experience with other health and safety laws, Congress realized that setting and meeting an exposure limit may not be sufficient to prevent disease. The Act, therefore, contains a detailed description of additional provisions to be included, where appropriate, in mandatory safety and health standards. One of these additional provisions requires that miners be informed about the nature of the hazards associated with their job and about the means for their own protection. Another provision requires the use of labels and other forms of warning to inform miners about the hazards, about proper precautions for safe use or exposure, about relevant symptoms of overexposure, and about emergency treatment. We have partially implemented these provisions through regulation. The Act requires also that standards shall prescribe protective equipment and control of technological procedures and that they shall provide for the monitoring of miners' exposures. We have implemented this provision to a degree also through regulation. Finally, the Act specifies that, where appropriate, a mandatory standard must prescribe the type and frequency of medical examinations in order to most effectively determine whether the health of miners is adversely affected by exposure. Additionally, when a determination is made that a miner may suffer material impairment of health or functional capacity by reason of exposure, that miner must be removed from such exposure and reassigned. In order to encourage miners to take the medical examination, the Act also provides that the miner shall not suffer loss of pay as a result of being reassigned. This additional provision on medical examinations has not been implemented by regulation for radiation hazards, nor for most other toxic substances or hazardous physical agents. We believe that this conference will provide us with valuable information on medical examinations for radiation exposure, as well as on other considerations to be used in future regulations. We still question whether our basic exposure standard of four working level months per year is adequate to protect a miner's health. This conference is indeed timely because both MSHA and NIOSH have been reviewing this issue for some time. I would like to reemphasize MSHA's commitment to education and training. MSHA strongly believes that miners should be informed fully about the real and potential hazards associated with their work. They should know the nature of the hazard and the means for protecting themselves from the hazard. The Act also emphasizes the need for training miners by requiring that new miners receive 40 hours of training before working underground and 24 hours of training before working on the surface. The Act also requires eight hours of refresher training annually. This training must include information on the relevant health hazards and the potential consequences of overexposure. Before I close, I would like to mention briefly another source of principles that MSHA often consults to obtain guidance for both the development and enforcement of regulations. These are the various court decisions that are handed down from time to time. Two in particular that I want to stress are the recent Supreme Court Decisions on benzene and cotton dust. Although neither of the decisions has a direct impact on mining or radiation, they do clarify some general principles and directions for regulatory agencies, such as the role of cost-benefit analysis and the need for risk assessment. In closing, I wish us all an open exchange of information for a productive and useful conference.

Jan 1, 1981

By J. E. Oberholtzer, M. G. Fernald

INTRODUCTION Approximately six years ago the U.S. Bureau of Mines (USBM) developed a data acquisition system (DAS) specifically designed for measuring radon levels and other environmental parameters during studies of means to control radiation hazards in underground uranium mines. The DAS system records data in machine readable form using a paper tape punch, which represented the state-of-the-art at that time for a moderate cost output device. However, the use of paper tape as a recording medium for field studies is somewhat unwieldy. Reducing the raw data required either that the tape be shipped to a computer center equipped with a high-speed paper tape reader or that the tape be transmitted at low speed over the telephone lines to a remote computer. Transmitting, at ten characters per second, the data from a 10-channel DAS taking Four readings per hour would require about 30 minutes For each 24-hour day's data. Telephone lines from remote mine sites are often of marginal quality and data errors can be introduced during transmission. Paper tape punches are also prone to occasional punching errors. Both problems make it necessary to carefully check for and correct data errors, a process which is possible because each DAS produces an independent printed data record, but the error checking and correction process can be quite laborious. Aware of recent advances in microcomputer technology which have brought the price of a personal computer down to about the cost of a paper tape punch 5-10 years ago, the Bureau decided to explore the feasibility of using a low-cost personal computer in the field to process DAS data in real time. On behalf of the Bureau, Arthur D. Little, Inc., developed a simple interface circuit which permits an Apple II computer to accept data from one or two DAS units as it is being transmitted to the paper tape punches. Computer software converts each measurement to appropriate engineering units, e.g., radon concentration, Working Levels, air velocity, temperature, or barometric pressure. The computer also calculates 1-hour and 8-hour running averages of all converted data and prints those results as soon as they are obtained on a line printer located at the test site for immediate inspection. After development, the system was used continuously and successfully for a 5-month period at a Utah uranium mine. DAS DESIGN AND MODIFICATION Each of the two USBM data acquisition systems used in this work consists of two separate modules. A multiplexer module located below ground near the measurement transducers acquires signals from each of nine tranducers. Six input channels were devoted to measurements of radon or Working Level. The outputs of those transducers, photomultiplier tubes or G-M tubes, respectively, are digital pulse trains which are accepted directly by the mutliplexer. Three channels were used for environmental parameters--air velocity, temperature, and/or barometric pressure. Each of the environmental tranducers is fitted with dedicated linearizing and voltage-to-frequency conversion circuitry so that the outputs to the multiplexer are also pulse trains having frequencies of one tenth of the value of the measured parameter expressed in the appropriate engineering units. A 100-Hz reference signal was input into the tenth channel for use in monitoring system integrity and performance. All ten pulse trains are then timeseries multiplexed into a signal line for transmission to the above-ground data acquisition module. Above ground, the composite signal is de-multiplexed into ten separate lines, each of which is connected to a digital counter which converts the pulse train to a numerical value. The acquisition of each set of readings is initiated by an adjustable "scan cycle comparator" timer. The acquisition process proceeds in three phases. First, radon and Working Level channels are counted for an extended period of time, typically 5-10 minutes depending on activity, because of the low pulse rates involved. Then the other four channels are counted for ten seconds, and finally, all ten readings, along with the Julian day and time of day are output serially onto paper tape and printed on a strip printer. When the scan cycle comparator reaches its preset time (15-minute cycle times were used in this work), it resets itself, initiates another readout cycle, and begins timing again. The only modification made to the data acquisition systems used in this work was to disconnect the scan cycle comparator in one unit, which became the "slave" and bring in the scan cycle comparator signal from the other unit, the "master", to initiate data acquisition cycles in the slave. Synchronizing the two data acquisitions in this fashion and using two slightly different radon counting times insured that the two systems never attempted to output data to the Apple II at the same time. THE APPLE II COMPUTER The Apple II computer used in this work was equipped with 48 KBytes of semiconductor random access memory (RAM), two floppy diskette drives, a Centronics Model 730 impact matrix printer and a modulator for driving an ordinary color television as a video display device. A single California Computer Systems Model 7720A dual 8-bit bidirectional parallel input/output (I/O) card was installed in the Apple to accept the digital data from both data acquisition systems. This card is

Jan 1, 1981

By Carroll F. Hardy

The day by day loss of industrial plants to gas and oil is chiefly by default. The coal industry is not selling its superior economy, safety, and other advantages to its customers. THE position of the coal industry has been affected by a wide variety of developments in the production and use of energy. The tempo of development and change has been increasing and the end is not in sight. Legislation is currently being proposed for commercial use of atomic power, and the employment of atomic energy in significant quantity will probably occur about the same time as the decline in production of petroleum and natural gas. But these developments are in the future and have little immediate effect on utilization and marketing of coal. While no one should try to suppress or retard the development of a new and economical source of energy, both the coal and private utility industry should be allowed to question how the nuclear power is to be used, who is to use it, and who is going to pay for it. The taxpayers have a monopoly on fissionable material and the knowledge to employ it. Any commercial use must stem from this source. It is not hard to visualize either taxpayer-subsidized private utility atomic power plants on one hand and super TVA's on the other. In view of the gains of gas and oil in the home heating field, it is interesting to compare the 1940 and 1950 census reports on the kind of fuel used for heating in occupied dwelling units. Table I shows that whereas coal provided 77 pct of the fuel for central heating (furnaces and boilers) in 1940, it was down to 45.4 pct in 1950. However, only about 1 1½ million units were lost in this 10-year period. In the non-central heating category, which principally includes stoves, the percentage declined from 39.2 to 25.6, but the units declined about 2½ million in number. The big increase was in heating units designed to burn gas and oil. Use of wood for central heating declined about one-third. Data on amount of fuels used for residential heating are not available, but information is on hand for residential and comnlercial space heating, see Table 11. Commercial space heating includes office buildings, churches, schools. and similar structures. The annual use of bituminous coal in these two categories declined about 1 million tons in the 10-year period. Other forms of solid fuel showed greater losses, except wood, which remained the same. Domestic stokers reached their high point in 1948 with about 1,200,000 in use. At the end of 1951 there were approximately 1,116,790 stokers in use. Conversions to gas and oil have been from hand-fired heating plants in the ratio of about 7 to 1 compared to stokers. In other words, for every one stoker which has been converted to gas or oil, seven hand-fired units have been converted to gas or oil. A bare recital of these data would indicate that the coal industry is holding its own reasonably well. However, 93.4 pct of the new homes built in 1951 were heated by gas or oil. Oil-burning equipment was installed in 37.8 pct and gas equipment in 55.6 pct of the new homes. This indicates that the public prefers gas when it is available, and that oil is second choice, with all forms of solid fuel apparently used when it is unavoidable. It must be pointed out, however, that during the period of rapid expansion of gas pipelines gas has been sold for house heating at prices that are in some cases actually lower than coal prices, or very nearly on a par. Gas has been sold at wells at far below the comparable price for oil produced from the same wells, and far below its actual worth. This situation is being remedied at the present time by increases in gas prices at the wells. For example, the wellhead price of gas in Texas averaged 7.494 per Mcf in 1952. In 1949 it was 4.59c per Mcf. This increase in price is being reflected in pipeline gas prices, and in most of the markets served by the pipelines the tendency is to get it out of the bargain basement type of sales. The American Gas Association estimates that at the end of 1952 there were in the United States about 11 million customers for gas house-heating, and the Association expects additional gains each year until around 18 million homes will be heated by gas in 1975. By 1975 there should be 60 million dwelling units to be heated in the United States, if dwelling units increase at the same rate as the population. If the gas industry heats 18 million dwelling units by that time, this still leaves 42 million units to be heated by some other fuel. If oil is used to heat 18 million dwelling units in that same year, 24 million would of necessity be heated by coal, coke, wood, electricity, or another fuel. The total number of dwelling units using coal listed in the 1950 Census was 18,776,000, so it would appear that coal has a chance at least to stand still in the tonnage sold for domestic use. In the first quarter of 1953, 2044 domestic stokers

Jan 1, 1955

By Brandon H. Grove

In Austria's only producing field, at Zistersdorf, Erdoel Produktions-gesellschaft completed a single new producer, its well Gosting 6 coming in early in October for an average production of about 15 tons daily. For further testing within the proved part of the field this company's Gosting 7 was at the close of the year drilling in Pannonian strata at 760 m., and its Gosting 8 was rigging up. Outside of the proved area of the Zistersdorf field two further tests located on the Steinberg dome failed to find production in the known oil-bearing strata of the Sarmatian, and were deepened as wildcats into the Cretaceous Flysch. The first of these, Neusiedel 1, about one kilometer north of the known producing area, found oil in the Flysch beds at a depth of 1290 m. for an initial production of about 10 tons per day. The second test, Windisch Baumgarten la, is an old well previously considered a failure, which when deepened achieved a production of about 5 tons per day in the Flysch Inoceramus beds at 985 m. Neither of the foregoing wells is producing from a stratigraphically definable horizon, oil accumulation in the Flysch of this area being apparently controlled by tectonic breccias or related crevices.

Jan 1, 1937

By Shiou-Chuan Sun

THE fact that coals of different ranks and even of the same rank differ greatly in their amenability to iroth flotation is well known. In recognition of the need for an explanation of this phenomenon, two hypotheses have been suggested. Wilkinsl reported that the floatability of coals increased with an increase of the carbon content or rank. This postulate is handicapped by the fact that bituminous coals that possess moderate carbon contents are actually more floatable than anthracite coals that have high carbon contents, as shown in columns 6 and 9 of Table I. Taggart and his associates' implied that the difference of floatability between bituminous and anthracite coal was caused by the variation of carbon-hydrogen ratio. This is not applicable to the relative floatability of other coals and carbons. For example, column 11 of Table I shows that the carbon-hydrogen ratios of low-floating lignitic coal and non-floating animal charcoal are not only smaller than the moderate-floating anthracite coal, but are also similar to the high-floating bituminous coal. Furthermore, according to this hypothesis, high temperature coke-A (464), Ceylon graphite (1238), and lamp-black (357), all possessing extremely high carbon-hydrogen ratios, should be less floatable than other substances having much lower carbon-hydrogen ratios such as high volatile-B bituminous coal (11.9 to 22), anthracite coal (35.7 to 60.5), lignitic coal (15.6 to 33.6), and charcoal (13 to 26.2). However the former group is actually more floatable than the latter group. In this paper, a surface components hypothesis is Proposed to explain the different floatabilities of coals, carbons, and hydrocarbon minerals. The validity of the hypothesis is experimentally supported by the actual floatability, natural floatability, wettability, and adsorbability for neutral oils of coals, carbons, and hydrocarbon minerals tested. The combustible recovery of the flotation results, as used in this paper. was calculated from Eq. 1: P (100-Ep) 100 RWCP Rc= [1] F (100-E,) C, where R, is the percent combustible recovery; F and P are, respectively, the weight of feed and the weight of concentrate or product; E, and Ep are, respectively, the total percent of ash plus moisture in feed and in concentrate; Ru. is the percent weight recovery: and C, and C, are, respectively, the percent of combustible in feed and in concentrate. Except for ash and moisture content, all chemical components of a coal are assumed combustible. The experimental work included studies on flotation, ultimate and proximate analyses, contact angle tests, extractions of bitumen-A with benzene, adsorptions for liquid hydrocarbons, and wetting tests. Most of the flotation experiments were performed in a laboratory Fagergren machine; others were tested in a small Denver machine. The solid feed for the former was 300 g and for the latter was 30 g. The solid materials used for flotation were crushed to —48 mesh. After the mineral pulp in the flotation cell was agitated for 6 min and the pH was adjusted to 7.5 & 0.2 with sodium hydroxide or hydrochloric acid, a petroleum light oil having a viscosity of 5.73 centipoises at 77 °F was added and conditioned for 2 min. Finally, pine oil was introduced and the froth was collected for exactly 3 min. The weight ratio of petroleum light oil to pine oil was kept constant at 1.5 to 1. Tap water was used for all flotation tests. Contact angles were measured with a captive bubble machine. For each coal sample, three specimens were mounted in transoptic mounts and polished with levigated alumina, first on a sheet glass, then on a cloth-covered metal polishing wheel. The polished specimen was first washed with distilled water and wiped thoroughly on a cleaned linen pad, then transferred into the pyrex cell of the captive bubble machine and conditioned for 6 min., and finally measured for contact angles at three or more points. Except where otherwise stated, the induction time for each measurement was 1 min. The contact angle representing each material was obtained by averaging the measurements of three specimens. The linen pad was first washed with warm distilled water, then boiled 30 min in a 2N sodium hydroxide solution, and finally washed with distilled water until no trace of sodium hydroxide could be detected in the decanted solution. The cleaned linen pad was stored under distilled water. Immediately before using, the pad was rewashed and transferred into a clean pyrex petri dish partly filled with distilled water. The glassware and rubber gloves used were cleaned by soaking in sulphuric acid-potassium dichromate cleaning solution, followed by rinsing with distilled water. The polished specimens were handled only by glass forceps. The ultimate and proximate analyses were made in accordance with the ASTM standard procedures for coal and coke. The extractable bitumen-A was determined by weighing 1 g of —100 mesh sample and placing it in a desiccated and weighed ASTM aluminum-extraction thimble. The thimble was placed in condenser hooks and inserted into an extraction flask containing 100 cu cm of benzene. The flask was heated and the benzene vapor was condensed by water coils. At the end of 24 hr of percolation, the thimble was removed, desiccated, and weighed. Loss in weight of sample was taken as bitumen-A and calculated to dry and ash-free basis.

Jan 1, 1955

By R. F. Hehemann, S. T. Zegler

The w phase in zirconium alloys containing more than G pct Nb can form in a difjUsionless manner during quenching or with composition change during aging at temperatures below 550°C. The latter treatment establishes a metastable equilibrium between an w phase containing 5 pct Nb and a ß phase containing 45 pct Nb. The superconducting transition temperat~ire of quenched Z?-10 pct Nb and Zr-15 pct Nb alloys is increased by aging at 375°C and both alloys exhibit the same Tc after prolonged aging. The superconducting transition both in the quenched and in the aged alloys is relatively sharp (occurring ovev a temperature range of <0.5°K), and is interpreted in terms of the microstructure that characterizes the ß-w) mixture and proximity effects postulated for thin-film superconductors. At elevated temperatures alloys of the transition elements from groups 4 to 6 in the periodic table exhibit wide ranges of mutual solubility in the bcc structure. Except for compositions that are high in a group 4 element? the bcc structure can be retained at room temperature by quenching and several systematic studies of the influence of composition on superconductivity in bcc alloys have been reported.&apos;&apos;2 The high-temperature bcc form (ß) of the group 4 elements—titanium, zirconium, hafnium—is not retained upon quenching but instead transforms marten-sitically to the hcp structure (a). The Ms temperature can be lowered by alloying the group 4 metals with a solute from a higher group and with a critical amount of solute the martensitic transformation can be suppressed. The 3 phase, however, is not retained without modification. Rather, a metastable phase, w (in particular in titanium- and zirconium-base alloys), occurs over a fairly wide range of compositions above that required for suppression of the martensitic transformation.3,4 As in the formation of a, the transformation of ß to w occurs in a diffusionless manner during quenching and the temperature for its initiation (ws) is lowered by increasing solute content. The ws temperature can be well below room temperature and reversibility of the transformation allows w that forms on cooling to low temperatures to disappear again when the sample is returned to room temperature.5 In Zr-Nb alloys the w phase can also be formed during aging at temperatures below 550°c.5 In this treatment, a metastable equilibrium is established between a niobium-rich ß phase and a zirconium-rich w phase. The present study of Zr-Nb alloys was undertaken to examine the influence of microstructure: as controlled by composition and heat treatment, on the superconducting transition temperature, Tc. MATERIALS AND PROCEDURE The alloys were prepared from zirconium crystal bar that was spectrochemically pure except for 30 ppm of Al, 40 ppm of Cu, 100 ppm of Fe, and 1 ppm of each B and Pb. The niobium used was 99.5+ pct pure containing according to the suppliers&apos; (Stauffer Chemical Co.) analysis a maximum of 50 ppmO, 60 ppm of N, 30 ppm of C, 5 ppm of H, 200 ppm of Ta, 50 ppm of Mo, and 15 ppm each of Fe, Co, and Ni. The alloys were prepared in the form of buttons weighing approximately 4 g by arc melting on a water-cooled copper hearth in an inert-gas atmosphere. All alloys were heat-treated in the ß-phase region for 3 days at 1000°C and water-quenched. For heat treatment, the specimens, 0.5 cm in diam by 1.0 cm long, were wrapped in zirconium foil and sealed in an inert-gas atmosphere inside quartz capsules. Quenching was done by breaking the capsules under water. Subsequent aging treatments of Zr-10 pct Nb and Zr-15 pct Nb alloys were done by placing the specimens in a lead bath at 375°C and quenching in chilled brine. The phases present in the quenched and aged alloys were identified by optical and X-ray metallography. X-ray powder patterns were taken at room temperature with a 114.6-mm-diam Debye-Scherrer camera and CuK-a radiation. Superconducting transition temperatures were determined from magnetic-permeability measurements6 in a 10 oe field. The superconducting transition temperature was determined from a plot of galvanometric deflection as a function of temperature by extrapolation of the linear portion of the curve to zero deflection. In Figs. 1 and 3 the arrows indicate the temperature range of the transitions. RESULTS AND DISCUSSION Superconductivity in Quenched Alloys. In Fig. 1 Tc for alloys quenched from 1000°C is plotted against nominal composition. Indicated also in Fig. 1 are results of metallographic studies. Alloys containing less than 5 pct Nb transform martensitically to a during quenching from 1000°C and Fig. 1 demonstrates that Tc increases significantly as the niobium concentration in a is increased. On the basis of the Bardeen-Cooper-Schreiffer theory of superconductivity7 it seems most likely that this increase in Tc results from an increase in the density of electronic states as zirconium is alloyed with niobium. This increased density of states has been revealed clearly in low-temperature specific-heat measurements.8 The martensitic transformation to a is suppressed when the alloy content exceeds 5 pct Nb. In the range from approximately 6 to 30 pct Nb, quenched alloys consist of a mixture of B and w phases and the amount

Jan 1, 1967

By R. F. Dickerson, D. A. Vaughan, A. F. Gerds

ALTHOUGH the metallurgy of uranium has been under intensive study since the early 1940&apos;s, no systematic effort has been made to identify the non-metallic inclusions in uranium. Uranium carbide (UC), which is probably the most common inclusion found in graphite-melted metal, has been tentatively identified by previous investigators, but the other nonmetallic inclusions have received little attention. Since metallography is a valuable tool in metallurgical studies, the metallographic identification of the nonmetallic inclusions in uranium is important. Such an investigation has been completed and the identification of slag-type inclusions and of uranium monocarbide, uranium hydride, uranium dioxide, uranium monoxide, and uranium mononitride is described. Metallographic Preporation It is often possible to prepare specimens for metal-lographic examination equally well by several methods. The specimens which were examined in this work were prepared by one of two acceptable methods. For the convenience of the reader, both methods will be discussed in detail and will be referred to simply as Method I or Method II in the subsequent sections. For both Methods I and 11, specimens for microscopic examination usually were mounted either in bakelite or in Paraplex room temperature mounting plastic. Method I—Specimens were ground in a spray of water on a revolving disk covered successively with 120-, 240-, and 600-grit silicon carbide papers. It was necessary to perform the final grinding operation carefully on worn 600-grit paper to keep the scratches as fine as possible. After washing and drying, the specimens were polished for 3 to 4 min on a slow speed wheel (250 rpm) covered with a medium nap cloth. Diamet Hyprez Blue diamond polishing paste, Grade 00, 0 to 2 µ, was used as abrasive with kerosene as lubricant on the wheel. Specimens were washed thoroughly in alcohol and final polished electrolytically in an electrolyte composed of 1 part stock solution (118 g CrO, dissolved in 100 cm3 H2O) with 4 parts of glacial acetic acid. A stainless steel cathode was used. At an open circuit potential of 40 v dc, a polishing time of 2 sec retained inclusions well with the bath at room temperature. If additional etching was required to sharpen the interface between the metal and the inclusions, an electrolyte composed of 1 part stock solution (100 g CrO3 and 100 cm8 H20) and 18 parts glacial acetic acid was used at room temperature. Best results were obtained by etching for from 10 to 15 sec at 20 v dc in the open circuit. Surfaces obtained by this method are suitable for microscopic examination. However, if desired, they may be etched further with other chemicals. Method 11—Rough grinding was done on a wet 180- or 240-grit continuous grinding belt. The specimen was then ground by hand successively on 240-, 400-, and 600-grit silicon carbide papers in a stream of water. Final polishing was accomplished on a 4 in. high speed wheel (3400 rpm) covered with Forstmann&apos;s cloth. Linde B levigated alumina, suspended in a 1 volume pet chromic acid solution, was the abrasive. Specimens usually were polished in 5 min or less by this technique. Often the inclusions present in the metal were identified in the mechanically polished condition. When etching was required to outline inclusions more sharply, one of the two following methods was used. In the first method, the specimen is etched lightly while electropolishing in the chromic-acetic acid solution described above (1 part of stock solution to 4 parts of acetic acid). The electrolyte was refrigerated in a dry ice-ethyl alcohol bath and specimens were etched at 60 v dc on the open circuit for 2 or 3 cycles of 3 to 4 sec each. The second technique utilizes electrolytical etching at about 10 v dc (open circuit) in a 10 pet citric acid solution at room temperature. X-Ray Diffraction Technique The major problem in the identification of inclusions in metals by X-ray diffraction techniques is the extraction of a sufficient amount of each type of inclusion to obtain an X-ray diffraction pattern. In the present study, X-ray diffraction patterns were obtained from individual inclusions of the order of 10 µ diam. The polished and etched samples shown in the micrographs were examined at a magnification of X54 or XI00 with a binocular microscope. This allowed sufficient working distance to extract the inclusions with a needle probe for powder X-ray diffraction analysis. Friable inclusions such as MgF2, CaF2, UO2, and UH3 could be freed from the metal by probing the as-polished and etched surface. The fine particles then were picked up on the end of a Vistanex-coated glass rod (0.002 in. diam) which was held in a brass adapter made to fit the powder X-ray diffraction camera. The end of the glass rod was centered in the path of the X-ray beam. In the case of the UC, UO, and UN inclusions which are smaller in size, more metallic in appearance, and less friable than the other inclusions, it was necessary to etch the inclusion in relief before extraction. UN inclusions etched sufficiently in relief in the electrolytic polishing solution described in Methods I and II by increasing the polishing time. UN inclusions were relief etched by extending the

Jan 1, 1957

By J. D. Livingston, H. E. Cline

The lamellar-dendritic transition in Sn-Pb alloys near the eutectic composition has been studied at high growth rates. Lamellar structures were found over a substantial range of tin-rich compositions, and this range extended to increasingly tin-rich concentrations as growth rate increased. These results are discussed in terms of stability and competitive-growth arguments. Various experimental and structural limitations to the rate of directional solidification are discussed. The rate of heat removal at the heat sink is the major experimental limitation. ReCENT interet1,2 in the use of fine composite structures produced by directional solidification of eutectic alloys makes it important to determine the range of composition and growth conditions that yield such microstructures. Because increasing growth velocities produce increasingly finer microstructures, it is particularly significant to determine the factors limiting the rate of solidification. Mollard and Flemings3 have shown that composite structures, free of primary dendrites, can be obtained in Sn-Pb alloys of off-eutectic composition. The composition range of composite structures was found to increase with increasing values of G/V, where G is the temperature gradient and V is the growth velocity. These results are in good quantitative agreement with an analysis of the stability of a planar eutectic interface.4 This analysis specifically predicts that over a small range of compositions stable lamellar structures will be obtained even for G/V = 0, hence, even at very high growth rates. The lamellar-dendritic transition in Sn-Pb alloys has also been analyzed with a model based on competitive growth between dendrites and the composite structure.576 This treatment, based on earlier work on organic eutetics,7 predicts that the composition range yielding composite structures in Sn-Pb will increase rapidly at high growth rates. An increase in the composition range of composite structures at high growth rates was recently observed in Cu-Pb alloys near the monotectic composition.8 In view of these results, and the predictions of the stability and competitivegrowth analyses, it was decided to study the lamellar-dendritic transition in Sn-Pb alloys at high growth rates. EXPERIMENTAL Using 99.999 pct pure materials, a series of Sn-Pb alloys were prepared containing 16.8 at. pct to 27.6 at. pct lead. (Eutectic composition is 26.1 at. pct Pb.) Ingots were extruded to 0.175 in. rod, and some rod was drawn to 0.070-in. wire. Directional solidification was accomplished in two different ways, Fig. 1. For growth rates up to 2 x 10-1 cm per sec, a 0.175 in. diam sample was placed in a graphite crucible 5 in. long with 0.250 in. OD and 0.035 in. walls. Samples were melted under flowing argon in a vertical, platinum-wound furnace, and solidified by driving the crucible downwards through a \ in. hole in a water-cooled copper toroid, Fig. l(a). An insulated chromel-alumel thermocouple was imbedded in the center of a representative sample, and moved with the sample during solidification. The local temperature is plotted against the distance travelled by the sample in Fig. 2. As the growth rate increased, the solid-liquid interface moved closer to the water-cooled toroid and the temperature gradient increased. At growth rates above 10-1&apos; cm per sec, heat was not removed fast enough and the sample moved into the toroid in the liquid state. The curve for V = 2 x 10-1 cm per sec shows a plateau caused by incomplete removal of latent heat from the interface, a problem which will be discussed later. To improve the heat removal, the toroid was cooled by nitrogen gas precooled in liquid nitrogen. This allowed successful solidification at rates up to 2 x 10-1cm per sec. Higher solidification rates required still more effective heat removal. Samples 0.070 in. in diam were placed in graphite tubes 0.125 in. in diam with 0.020 in. walls. Instead of cooling by sliding contact with a cooled toroid, these thinner samples were sprayed or directly immersed into water, Fig. l(b). After solidification, samples were stored in liquid nitrogen until they could be examined metallographic-ally. The surface was prepared with a diamond-knife microtome, followed by a light etch. The presence or absence of tin dendrites, Fig. 3, or lead dendrites, Fig. 4, was noted by optical microscopy, usually of a transverse section near the center of the sample. Replicas of the surface were prepared and examined in an electron microscope to resolve the fine lamellar structures, Fig. 5. The structures observed at various compositions and growth rates are summarized in Fig. 6. Composite structures were observed at increasingly cin-rich compositions as growth rate increased. This transition from dendritic to composite structure with increasing growth rate was also demonstrated by solidifying half a sample at a slow rate and then suddenly increasing the growth rate by lifting the furnace and quenching the sample with a water spray. A longitudinal section of this sample, Fig. 7, shows that the tin dendrites, which extended ahead of the slow-moving composite interface, were bypassed by the composite when the growth rate was increased. The range of composite structures at high growth rates was limited by the appearance of primary lead dendrites on the tin-rich side of the eutectic composition. Observation of representative longitudinal

Jan 1, 1970