Search Documents

By A. L. Field

Previous Publications THE first report on the slag viscosity work of the Bureau of Mines was made by one of the authors 1 in 1916. It was concerned chiefly with, the method of measurement. A paper 2 on this phase of the work was presented to the Institute in February, 1917, and a similar paper 3 was communicated to the Faraday Society. Purpose and Scope of the Slag Viscosity Tables Following the development of a suitable method, the next step in the investigation was the measurement of slag viscosity over a wide range of temperature and composition, and the arrangement of this experimental data in the form of tables for the use of the blast-furnace operator. It is believed that this information, if used intelligently, should help the operator to reduce losses in production caused by off-grade pig iron, to improve fuel economy, to promote operating efficiency, and to extend present-day practice to meet the increasing need of smelting lean and complex ores. In the experimental part of the work it was early observed that the temperature-viscosity relations of a slag were determined, for practical purposes, by the relative proportions of the three major constituents, lime (CaO), alumina (Al203), and silica (Si02), provided the percentage

Jan 12, 1917

By Earl Oliver

IN times like these, the A. I. M. E. and similar societies have their greatest usefulness. . . . Individuals and companies acting alone in the development of public opinion are merely voices crying in the wilderness. United and organized they make a great force that can and does influence public action. . . . What part, if any, then, shall the A. I. M. E. take in these great stabilization processes of civilization? Shall it help to give direction and force to scientific knowledge which comes within its field of activity or shall it merely develop more of it? . . . Is it the A. I. M. E.'s function to help bring system, order and direction out of this situation so that the world can again have forward economic movement, or should it rest on its oars until other agenciess develop that public understanding and direction so necessary for economic progress? As we understand the functions of the A. I. M. E. Petro¬leum Division, they are to discover and promote knowl¬edge for the benefit of society on the mining of oil and gas. To the individual member of the Petroleum Division, there are certain advantages of social and professional contacts, but the collective purpose is that described. And the individual is benefited by membership only as the Division measures up to the purposes of its existence. Consequently, the welfare of this Division and that of its members depends on the extent to which it keeps pace with the needs of society on matters relating to the mining of oil and gas.

Jan 1, 1932

Tunnel damage from explosion and earthquake excitation is correlated with strain and particle motions. Critical strains and motions are then correlated with focal distance, which is so small that permanent strains are highly likely. The shaking process is then numerically modeled to investigate the effects of excitation frequency and joint orientation. The models confirm that near cavern blocks slide intermittently only at relatively high particle velocities during periods of low joint normal stress. Larger displacements occur with relatively small shear wave lengths or relatively high frequencies. Variation in joint orientation and rock mass dilation result in changes in displacement.

Jan 1, 1985

By M. W. Mikhail, R. J. Mikula

"Funding has been made available through the National Energy Program to build a mobile fine coal treatment plant. The plant is intended f or field application and also to upgrade the existing pilot plant facilities.Western Canadian coals with high proportions of fines (due 10 their friability) cause high losses of fine coal to the reject and in turn reduce the over-all recovery of saleable coal. Eastern Canadian coals with high amounts of sulphur of ten require crushing to fin e sizes to liberate pyrite and therefore need effective and efficient separation of these fine sulphur particles. The difficulties involved in solving these problems are two-fold. One is the reluctance to experiment in an operating wash plant and the other is that laboratory scale testing often cannot be extrapolated to the plant environment. The mobile fine cleaning plant with its ability to operate at existing coal preparation plants across Canada and with its flexibility in implementing flowsheet option s and close monitoring of fine coal circuit variables, is ideally suited to solve these fine coal cleaning problems. The mobile plant objectives are to design and optimize flowsheets, evaluate and improve fine cleaning circuit operations and 10 lest new process monitoring and control equipment.The schedule for construction and the required resources will be outlined along with flowsheet options and the general R&D philosophy of the mobile plant.IntroductionIn relation to coal preparation technology, Canadian coal s can be grouped as Eastern coal, Rocky Mountain metallurgical coal and Foothills high volatile bituminous coal. The eastern coals have high sulphur contents and in some cases require crushing or even grinding to liberate the fine pyrite . Rocky Mountain metallurgical coals often occur in highly sheared seams which have been subjected to rapid coalification. The high friability of these coals means that degradation during mining, handling and processing is extensive and results in large amounts of fine s ranging between 20 and 60% by mass of the wash plant feed. The widely varying geo logical structure and exposure to the elements results in partially oxidized coals which perform inconsistently in the traditional method of fine cleaning by flotation . Foothills high volatile bituminous coals for thermal use are frequently interbedded with soft shales and bentonitic clays. These clays disintegrate during washing and swell, producing large volumes of tailings. The presence of clays in plant circuits, particularly fine circuits, tends to increase the viscosity and may cause deterioration of the fines density separation and increase magnetite losses. In view of the fact that fine coal cleaning is the most difficult aspect of coal preparation , a research program is being established to study the factors involved and to offer some solutions.A recent performance evaluation of Canadian washeries showed that yield losses in the minus 2 mm coal circuits were approximately ten times those occurring in the coarse coal circuits(1). Because 2 mm x 0 coal represents 30 to 70 per cent of the raw coal, these losses can be considerable . The operating and capital costs of beneficiating minus 6 mm coal are about four to five times higher than for plus 6 mm coal. In the case o f minus 2 mm coal the costs are even higher. The economic importance of the fine coal cleaning cannot be over emphasized and as a result, the CANMET R&D program is geared for fine coal beneficiation."

Jan 1, 1984

By Earl Oliver

This paper is based on the formula that the petroleum industry is sick—find the cause, eliminate it, and the industry will get well. There is perhaps no difference of opinion that the cause is overproduction, actual or potential. But that, again, is a result rather than a cause. The problem is to study the influences that serve to keep production and consumption out of balance. Production control is a popular theme and a natural one at this time when the whole world is unstable; but production control in its popular sense means little more than the application of emergency measures while danger threatens. Economists differ as to the effectiveness of these, most of which are designed to alleviate and not to cure, but we assume there are none who question the wisdom of promoting conditions under which economic forces are permitted to operate freely in bringing about stable conditions. The petroleum industry suffers from two classes of economic ills that serve to promote overproduction; namely, those which are common to all industries and those which are peculiar to this industry. It is of the latter I wish to speak. Strange to say, government is responsible in large part for this latter class of ills that affect the petroleum industry and we are coming to recognize that only through the assistance of government can they be corrected. Instability of the petroleunl industry, in so far as human agency was involved, had its inception in 1875 when a Pennsylvania jurist, called upon to make a decision for which there was no precedent and fancying he noted some resemblance between the so-called fugitive nature of oil and gas and that of wild game, made mention of that resemblance. Other courts being impressed by the analogy enlarged upon it, and gradually fixed upon oil and gas a legal status in certain important respects similar to that of wild game. Although later the courts abandoned this analogy, certain practices that developed out of that early concept have continued to this day and are largely responsible for the industry's present plight. Thus, instead of ownership in place that is characteristic of private property generally, the law of capture was applied to oil and gas. Now, the rapid and wasteful depletion of game

Jan 1, 1932

By Earle J. Klohn

INTRODUCTION The paper presents a broad overview of the geotechnical site studies required for siting uranium tailings dams. These studies, which are interdisciplinary in nature, touch upon such related fields as: hydrogeology, hydrology, and seismicity. Tailings dams are commonly used to develop tailings ponds for the storage of the waste products generated by mine milling operations. In the past, most tailings ponds have been developed either above ground on dry land, or in existing lakes. In the United States a majority of the uranium mines are located in arid or semi-arid regions and past practice has been to store the wastes above ground on dry land (1,2). Currently, the United States Nuclear Regulatory Commission (N.R.C.) recommends below ground disposal of uranium waste products, although they indicate that they will consider above ground alternatives (3). In Canada many of the existing uranium mines are located in a wet environment and past practice has been to use small lakes and depressions as tailings ponds (1,4). However, current practice in Canada leans towards above ground, dry-land storage (5,6,7), and the Atomic Energy Control Board (AECB) guidelines (5) indicate this to be the preferable method. In B.C., where a moratorium on uranium mining has been declared by the Provincial Government, there are no existing uranium tailings ponds. However, both the formerly proposed Birch Island and Blizzard uranium developments would have used dry land storage. Most of the existing non-uranium mill wastes in B.C. are stored behind tailings dams on dry land. The waste products delivered to the tailings pond comprise, the ground-up rock that remains after most of the uranium values have been removed and the excess chemicals used in the milling process. These waste products, which are called "tailings", are normally mixed with water to produce a slurry and then pumped to the tailings pond. Since the uranium recovered constitutes only a small percentage of the ore (usually less than 1%) (8), large quantities of waste rock and associated transportation water are discharged into the tailings impoundment area. Consequently, relatively large tailings pond areas are required for the uranium tailings. In recent years, major advances have been made in the design, construction, and operation of non-uranium tailings storage facilities (9,10,11,12,13,14, 15). These improvements have been based on combining the latest advances

Jan 1, 1980

By Eugene J. Schreiber, F. Newton Parks

Booz-Allen & Hamilton management consultants has recently completed a comprehensive study of Brazil's minerals and metals sectors for 40 leading international and Brazilian companies. The 220-page study describes the activities and Investments under way in 25 key commodities and assesses the current state of Brazil's economy, the impact on domestic minerals and metals industries, and government minerals policies.

Jan 12, 1977

By T. W. Oshnock

Pulverized coal (PC) and other injectants can be injected into the blowpipes of a blast furnace to decrease coke requirements. Throughout the history of blast-furnace development, there have been many attempts at injecting PC. In the 1960's a more serious effort was made to improve pulverized coal injection (PCI) technology. However, because of the additional capital requirements and the projected near-term inexpensive cost of other injectants, very few PC1 systems were installed. When the second oil crisis occurred around 1979, the shortage and high cost of fuel oil caused blast-furnace operators to operate without oil injection. It was soon discovered that blast- furnace operation without oil was not as smooth or efficient. Experiments with PC1 quickly showed that coal was a good replacement for oil which resulted in the rapid application of PC1 in both Europe and Japan. As a result, PC1 technology for blast-furnace operation has progressed rapidly during the 1980's. This paper will describe the worldwide status of PCI, characteristics of different PC1 systems, the effect of PC1 on blast-furnace process parameters, and the economic effects of using PCI. Results from the USS Blast-Furnace Thermochemical Projection Model comparing the effects of different coal properties and the effects of other tuyere injectants such as oil and natural gas will also be described.

Jan 1, 1988

By Dennis E. Stover

Grade control for in situ uranium leaching is maintaining, at desired levels, the uranium concentration in the pregnant lixiviant which feeds the extraction (ion exchange) circuit. This differs from grade control in conventional mining in that in situ grade control is imposed subsequent to rather than prior to leaching without the direct observation and control possible in conventional mining and milling. Grade control is a necessary engineering task through the entire project life, beginning with preliminary evaluation and design and ending only when the final wells are shut in. More specifically, this task is associated with designing the initial well field, maintaining productivity during well field operations, and planning well field expansions to compensate for the depletion of existing wells. Our aim during each of these phases is to maximize uranium productivity while ensuring that we are developing information necessary for future phases. Throughout this task, we remain cognizant of the fact that uranium productivity is a function of both the complex geologic and operational parameters of the project. A minimum list of significant geological factors must include the following: . Host formation mineralogy . Uranium mineralogy . Permeability . Ore quality (grade and thickness) Our concerns are not limited to average or general descriptions of these factors. The interactions among the factors as well as localized variations among them are far more important in obtaining high productivity from a deposit. For example, an ore body may display an average permeability of one darcy but have its high grade ore confined to regions with permeabilities less than 500 millidarcies. This relationship between ore grade and permeability must be understood early in the project life if the operation is to be successful. Similarly, the fundamental operating parameters for the well field include those items which enable us to distribute the lixiviant within the ore as well as the chemistry which will dissolve the uranium minerals. A minimum list of such factors will include: . Well design . Well efficiency . Well bore damage . Pattern size and shape . Oxidant type and concentration . Complexing agent concentration - pH The interactions between these variables and the complex geological variations create a situation which often defies quantitative analysis. Hence, we have adopted a statistical approach when correlating production history with these parameters and the quality of the correlations improves as the number of wells operated increases. However, the correlations are often specific to an ore body. We find that we must tailor the process to meet the unique geologic parameters of each ore body. PHASE 1: INITIAL WELL FIELD DESIGN Designing the initial well field is perhaps the most difficult phase because we have the smallest data base from which to predict performance. We will have conducted numerous laboratory tests and, perhaps, field pilot tests to select the lixiviant system and to characterize the geologic parameters. However, we will learn a great deal more about the deposit and its productivity during actual commercial scale operation. Fully acknowledging its limitations, we will, as a starting point, assume that uranium productivity will be directly proportional to the quality and quantity of recoverable reserves. Future deviations from this assumption will be plentiful and result from our limited understanding of the aforementioned interactions between and among the operational and geological factors. However, in planning the initial field, estimation of the quantity of leachable uranium present at levels above the economic grade and grade-thickness limitations is a priority task. (The concentration of uranium in the ore, the ore grade, is expressed in units of weight percent. The grade-thickness of the ore is the product of the average grade and ore thickness expressed in units of percent-feet.) These parameters provide a direct indication of the quality and quantity of ore. They are determined from gamma logging of closely spaced drill holes (50 to 100 foot centers being typical) accompanied by uranium core analysis or direct uranium assay logging. Gamma- logging measures gross gamma radiation which emanates from uranium decay products and indicates only the possible presence of uranium. Direct uranium assays are necessary to confirm this presence and the true location and quality of the reserve. From this information, an estimate of the in-place uranium reserve is formulated. Utilizing the above data, we begin to assess the quantity and quality of the leachable reserve. Analysis of self-potential and resistivity logs in conjunction with uranium logs will identify those zones which contain uranium and display apparent high permeabilities. Similarly, zones with extensive calcite cementation or dense clay sedimentation are located and uranium within such zones is excluded. (Interpretation of these logs should be confirmed

Jan 1, 1980

By Arthur E. Lord, W. Martin McCabe, Robert M. Koerner

Following the driving of each new section of tunnel, a prompt decision is required regarding the type and extent of primary support to control convergence and prevent ground failure. Sometime thereafter, the effectiveness of the in-place support must be evaluated over the entire length of excavation. It is therefore extremely desirable to have available a lightweight, portable, easy to install and interpret, cost effective monitoring system to assist with these ground control functions. Such a system can be assembled based on the acoustic emission (or microseismic) technique. This monitoring technique uses a piezoelectric transducer to detect subaudible noises associated with deformation of and/or crack propagation in the soil or rock being monitored. After detection, the signals are then amplified, filtered and electronically counted to provide a quantified assessment of the stability of the structure. In the simplest sense, the absence of emission, or counts, signifies no deformation and thus stability. Conversely, the occurrance of counts signifies instability to varying degrees. The higher the count rate, the greater the instability and the need for immediate or additional support. Described in this paper is an overview of the acoustic emission (microseismic) method, details of its use in a laboratory study measuring convergence and stand-up time in soil, and the behavior of rock samples under similar stress conditions. Field use of the method is recommended for ground control monitoring in tunnels, and some suggestions regarding monitoring procedure are presented.

Jan 1, 1979

I HIT the ceiling when I opened my November MINING ENGINEERING and after waiting a week to cool off I' still find it necessary to express my resentment of the gross libel embodied in the editorial introduction to Norman Stines' legitimate criticism of my paper on "Russia's Mineral Potential" (p. 949). The underlying thesis of my article, and a point that I tried to make even more clearly in my oral presentation at the St. Louis meeting last February, is that we should not over estimate the present and near-future mineral production of the USSR. The words that I resent in your boxed introduction are: "Opinions range from those presented by Paul Tyler ... to those of Norman C. Stines . . ." This is not merely an understatement that can be defended by admitting that the range goes much farther to the left, beyond me. It represents the appraisal that 90 pct of your members will carry away, namely that I am not only an incompetent observer-which may be true-but always that my bias is to applaud the efficiency of the beasts in the Kremlin-which is not true. Notwithstanding my efforts to clarify the record I now find myself branded as being at the opposite pole from 'Mr. Stines among the red-eyed worshipers of Communism or at least among the timid souls that shake in their boots in the belief that the Soviet system has created an invincible as well as an evil menace to the decent peoples of the world. My ideas are somewhere in the middle ground between Mr. Stines' complaisance or contempt and the exaggerated fears of several other observers whose views have received wide acceptance. PAUL M. TYLER MINERAL TECHNOLOGIST & ECONOMIST 5005 EDGEMOOR LANE, BETHESDA, MD. The introductory paragraph to Mr. Stines' criticism of Mr. Tyler's article (Russia's Mineral Potential, June issue MINING ENGINEERING, page 494) is erroneous as pointed out in author Tyler's letter above. In his article he objectively presented an estimate of the mineral position of the Soviet based on the statistics available to him which sources are credited in the first paragraph of his article. We happen to know that, Mr. Tyler was reluctant to prepare the article for the Mineral Economics Division sessions at the 1951 Annual Meeting and did so only because of his loyalty to AIME. It was thought that this article would be a basis for discussion of this important subject at the Annual Meeting. We apologize to Mr. Tyler for any embarrassment our introduction, may have caused him. Editor

Jan 1, 1952

By Rob Jameson

Micropiles are ideally suited to seismic retrofit applications. Their high tensile and compressive capacity can sustain the large concentrated loads induced during seismic events. Micropiles can be drilled through and grouted into the wide variety of soil and rock materials encountered in seismically active zones, using equipment that operates effectively within existing structures. Micropiles have seen increased acceptance and are now widely specified for retrofit work in the San Francisco Bay Area. While installation equipment and the structural configuration of micropiles have been largely unchanged since the early 1990?s, market forces have pressured both engineers and contractors to continually optimize the cost and performance of seismic upgrades. Instead of the contractor simply developing stronger, stiffer and cheaper micropiles as isolated structural elements in response to a specification, focus has shifted to maximizing the efficiency of micropiles within the overall foundation system. The case history of seismic retrofit project in Berkeley, CA is presented to illustrate the collaboration between designers and drilling specialist which resulted in a structurally efficient and cost effective design. The seismic retrofit program for this 6-story concrete building was developed through careful consideration of the geotechnical and structural mechanisms of resistance, in combination with the deformation response of high capacity micropiles (2670 kN) under ultimate loading conditions. This case history describes the interactive project development emphasizing performance-based design and field testing to optimize the micropile foundations and the overall structural system.

Jan 1, 2008

By D. Embury

The construction and utilization of oil and gas pipelines is a vital aspect of the Canadian economy. The evolution of materials for this application can be considered as part of an evolutionary process in which the processing of the materials is linked to the design requirements of the pipeline. The evolution of both materials and designs will be considered in terms of parameters such as yield strength, work hardening characteristics including strain path changes such as the Bauschinger Effect and toughness. The final portion of the presentation will consider the possible uses of non-metallic materials, composites and 'smart 'systems in pipeline construction.

Jan 1, 2005

By T. K. Smith

The comments contained in the discussion of the paper concerning the prediction of fracture gradients are of considerable interest, since the subject is of a very controversial and timely nature. It is regrettable and Fairhurst2 contradicts the author's statement in the third paragraph of Other Applications. These references create significant doubt that Eaton's Eq. 1 predicts fracture pressure during the injection of fluid into old wells. 3. No evidence is presented in this paper to support the author's Conclusions 2 and 3. While an empirical coefficient related to back-calculated field data is correlated with depth (Eaton's Fig. 4), there is no reason to suspect that this coefficient might be the Poisson's ratio of the various formations involved. 4. Until the field data supporting the empirical approach proposed by the author are presented and evaluated, the accuracy claimed in Conclusion 5 is without substantiation. References 1. Geertsma, J.: "Problems of Rock Mechanics in Petroleum Production Engineering", Proc., First Intl. Symposium on Rock Mechanics, Lisbon (1966) 585. 2. Haimson, B. and Fairhurst, C.: "Hydraulic Fracturing in Porous-Permeable Materials", I. Pet Tech. (July, 1969) 811-817. that the comments are those of a critic and do not offer an alternative and better method for the prediction of fracture gradients. Nevertheless, a reply is in order.

Jan 1, 1970

By P. Anyon

Exposure to the microscopic particles in diesel engine exhaust can lead to serious health problems, including the incidence of cancers, heart disease and increased susceptibility to respiratory ailments such as pneumonia, bronchitis, and asthma. In Australia, mining companies and state government authorities are actively responding to this major occupational health and safety issue. Their initiatives place considerable emphasis on routine monitoring of diesel particle matter (DPM) exhaust emissions, both for engine benchmarking and as a part of routine engine maintenance. Implementation strategies follow two complementary paths: determination of individual engines? emission signatures combined with periodic engine re-certification; and co-operatively agreed industry-government engine maintenance programs, which are subject to audit. Both strategies place heavy reliance on real-time measurement of DPM concentrations in engine exhaust, using ruggedized instrumentation, mainly based on laser light scattering photometry. Adding value to this cooperative approach, industry has also collectively funded a government agency to design and manage a database, into which the results of all-routine DPM testing can be uploaded directly by individual mines. This win-win outcome is a great example of how cooperative action between governments and industry can deliver the goods: in this case, a safer working environment, more efficient energy use and the capacity to benchmark performance with industry peers.

Jan 1, 2009

The great problem at Port Pirie and many other lead smelters has always been to keep the lead content of the slags down to a low level. This has been done in the past largely by empirical methods, the lead content varying from less than 1 per cent to over 4 per cent, the more usual figure being something under 2 per cent. Subsidiary to this, but of major importance in the future, was the extraction of the considerable quantities of zinc in the slag. At this stage little was known about the constitution of the slags apartfrom their chemical composition. Some early work on slag constitution was carried out in the B.H.A.S. Research Laboratories between 1919 and 1921, but except for a microscopic examination of a few thin sections the greater part of this investigation wasdevoted to the recovery of the zinc content of the slag by leaching methods.In 1949, therefore, it was decided to initiate a programme of fundamental research into the constitution and properties of the lead blast furnace slags. The information thus gained would give a clearer picture of reactions taking place in the slag, both in the blast furnace and during dezincing.There was also the hope that these results would lead to improved industrial techniques resulting in smaller lead losses, and, taking the long-term view, to efficient methods of producing zinc from the slags.A series of investigations was therefore begun at various centres over a period of time. These were:1. A mineralogical examination of the slags by Dr. A. B. Edwards of the Mineragraphic Section, C.S.I.R.O.2. A study of the thermodynamics of various slag com...

Jan 1, 1956

By Morris A. E

The thermodynamically-based pyrometallurgical simulation software STEPSOL (Version 3.0) provides for the modelling of activities using the Interaction Parameter Model, the Regular Solution Model, the Sub-Regular Three Suffix Margules Model, the Sub-Regular Four Suffix Margules Model and various simpler polynomial models. These are briefly reviewed, along with the major sources of data available in the literature, and the implementation of the models in the program.

Jan 1, 1991

By Morris A. E

The thermodynamically-based pyrometallurgical simulation software STEPSOL (Version 3.0) provides for the modelling of activities using the Interaction Parameter Model, the Regular Solution Model, the Sub-Regular Three Suffix Margules Model, the Sub-Regular Four Suffix Margules Model and various simpler polynomial models. These are briefly reviewed, along with the major sources of data available in the literature, and the implementation of the models in the program.

Jan 1, 1992

By S. K. Oraee

Most management decisions at all levels of the organization are as directly or indirectly depends on the circumstance of future. With regard to predict the future events in the process of decision-making plays a main role, therefore, forecasting is very important for every organizations and institutions. There is a variety of methods to predict time series. In general, these techniques can be divided as following: statistical, artificial intelligence and analytical techniques. Two of the most common methods for time series prediction is autoregressive integrated moving average (ARIMA) and artificial neural network (ANN) methods, these methods are the subset of statistical and artificial intelligence techniques respectively. In this paper, a hybrid model of ARIMA and ANN models are employed to predict the number of fatal injuries in the USA underground coal mines. This research showed the result of hybrid model is better than split model.

Jan 1, 2011

A. W. TRACY' AND D. H. THOMPSON (presented in written form).-Messrs. Ed¬munds, Anderson and Waring are to be commended for their excellent work in so painstakingly reducing to constants, the variables believed to be associated with these types of stress corrosion. An ammonia stress-cracking apparatus similar in many respects to the one described has been assembled. Certain aspects of its construction may be of interest. Tensile loading of the specimens, which are usually Y8-in. diameter wires, is accomplished by springs acting in compression in compact, portable, tubular assemblies weighing less than 10 lb. The lower end of each assembly contains the test specimen and is passed into the test chamber for exposure to the moist ammoniacal atmosphere. There is considerable convenience in being able to load the specimens in a portable unit, which may be warmed to prevent condensation, and then thrust instantaneously into the test chamber. Spring loading obviates the use of weights, but it has the disadvantage of allowing some relaxation of the load if the specimen elongates during corrosion. The maximum loss in load found so far was 20 per cent. The air supplied to the test chamber is humidified by passing it through thermo¬statically controlled water. The air is dispersed into fine bubbles by being forced through a porous ceramic filter at the bottom of a tank containing about 10 liters of water. The temperature of the water is controlled independently of the tempera¬ture of the test chamber. Thus, with the

Jan 1, 1945