Search Documents

By John S. Berber

Bureau of Mines research on the production of phthalic anhydride by vapor phase oxidation of hydrocarbons began at Morgantown, W. Va., in October 1960. For these preliminary studies, a known solution containing naphthalene and 1-methylnaphthalene was used. In another series of tests, pure naphthalene was dissolved in different solvents. A series of investigations was conducted with the neutral oil fraction (210° to 270° C) of a low-temperature tar obtained from the Texas Power and Light Company's low-temperature carbonization unit. The neutral oil fraction was first dealkylated, and then the product was distilled into 210° to 250° C and 250° to 270° C fractions. These two fractions were then subjected to catalytic vapor phase oxidation. Although a number of catalysts was tested, most of the experiments were made using 9 percent vanadium pentoxide on inert alumina. The activity of this catalyst did not decrease after 162 hours at 1,030° F. Studies were conducted over a temperature range from 610° to 1,180° F. An operating temperature of 1,030° F yielded 47.50 percent of phthalic anhydride from a synthetic mixture of naphthalene and 1-methylnaphthalene. The highest yield (84.40 percent phthalic anhydride) was obtained with naphthalene dissolved in toluene and also with naphthalene dissolved in methyl ethyl ketone. With a 210° to 250° C fraction the highest yield of phthalic anhydride obtained was 12.32 percent at a temperature of 900° F, and, with a 250° to 270° C fraction, it was 15.80 percent at a temperature of 875° F.

Jan 1, 1963

By James S. Browning

Recovery of fine mica from a sample of impounded tailing from the Dixie Mines, Inc., crushing and screening operations, Heflin, Ala., was investigated to determine if plant losses of mica could be reduced while econmically producing a commercially acceptable mica concentrate. Conventional wet-tabling and agglomerate-tabling methods yielded concentrates assaying about 95 percent mica from the tailing that contained 17 percent mica. Products containing about 98 percent mica and accounting for mica recoveries of 86 percent were obtained by flotation, using as reagents sulfuric acid, kerosene, and coco amine acetate (a mixture consisting chiefly of dodecyl and tetradecyl amine acetates). The results of the batch flotation tests were verified by tests in a small continuous flotation pilot plant.

Jan 1, 1963

By H. E. Powell

A laboratory investigation was conducted by the Bureau of Mines to study the magnetic characteristics and to measure the magnetic susceptibility of the mineral siderite. All susceptibility measurements were made by the method of Gouy on powdered samples and at room temperature. Specific susceptibilities of samples from 15 different locations, domestic and foreign, were determined. Values ranged from 66 to 103 (10-6) cgs electromagnetic units. The relation of susceptibility to mineral composition is discussed in this report. The specific susceptibilities of 76 minerals are tabulated in an appendix.

Jan 1, 1963

By F. B. Petermann

Two-stage electric smelting tests, using small submerged arc furnaces, were made on high-iron manganiferous materials from the Pioche District, Nev., and the Cuyuna Range, Minn. The object was to reduce sufficient iron in the first stage to give a slag with a favorable Mn to Fe ratio for smelting to ferromanganese in the second stage. The nodulized flotation concentrate from the Pioche District contained 29.2 percent Mn and had a Mn to Fe ratio of 2.3:l. When smelted with coke and silica or alumina as flux, more than 90 percent of the manganese was recovered in slags with Mn to Fe ratios of 10: :24:1. The Pioche slag was smelted to ferromanganese using coke and lime- stone or dolomite as flux. The ferromanganese was high in silicon but other- wise met specifications. Recovery of manganese to ferromanganese in this step was 73 to 78 percent. Overall recovery of manganese was about 70 percent. The sintered flotation concentrate from the Cuyuna Range material contained 12.9 percent Mn and had a Mn to Fe ratio of 0.3:1. When smelted with limestone as flux, about 83 percent of the manganese was recovered in a slag with a Mn to Fe ratio of about 6:1. The Cuyuna slag was smelted to ferromanganese with coke and limestone for flux. Recovery in this step was 61 percent of the manganese in a high silicon ferromanganese. Overall recovery of manganese was about 50 percent.

Jan 1, 1963

By D. G. Kesterke

The electrolytic reduction of uranium dioxide to molten uranium in fluoride baths at approximately 1,200° C was demonstrated. Electrolyte constituents were BaF2, UF4, and either MgF2 or LiF. Preliminary laboratory-scale experiments were made in which the uranium oxide was added continuously for 3 to 5 hours to the cell during electrolysis. Over 1,000 grams of metal of 99-8 percent purity was made in some tests, Major impurities were aluminum, barium, iron, silicon, carbon, and oxygen. The effects of cell design, geometry, and bath composition were investigated. Electrode arrangement was the most important factor in cell performance.

Jan 1, 1963

By Albert W. Deurbrouck

This report describes the preparation characteristics of the significant, presently producing coalbeds in Mercer County, W. Va. Approximately 650,000 tons of coal was produced in Mercer County, W. Va., in 1960, of which over 630,000 tons was cleaned in four preparation plants. In 1960 the Pocahontas Nos. 3, 6, and 11 coalbeds supplied 99.5 percent of the total coal mined in the county. Because the number of working mines In this county was small, only six samples were collected; four of which were of metallurgical quality as mined. The remaining two samples could be upgraded by washing at suitable specific gravities. Many of the samples taken were of metallurgical quality as mined; however, the float-and-sink analysis on these coals has been included to point out their amenability to the production of premium quality, special purpose coals. In general the coals were low-volatile bituminous in rank with low ash and sulfur contents.

Jan 1, 1963

By John L. Miller

The synthetic fluormicas may be divided on the basis of silicon content into tetrasilicic, trisilicic, and disilicic. Properties of lead and barium disilicic fluormicas were determined to characterize these two representative members of the disilicic group. Properties described include unit-cell parameters from X-ray diffraction, optical data, chemical composition, density, melting point, and mechanical and dielectric properties. Both the lead and the barium micas were typically monoclinic; neither was sufficiently cleavable to yield usable mica splittings. Both yielded polycrystalline, machinable ceramics with excellent mechanical and dielectric properties.

Jan 1, 1963

By Frank A. Peters

A series of evaluations is being made to determine the relative processing costs of producing metallurgical-grade alumina from low-grade aluminous materials. These evaluations include the estimated operating and capital costs as well as recommendations for additional research and development to improve the processes and to lower their costs. The design for each process is based on published work. The three processes considered in this evaluation are designed to leach alumina as aluminum sulfate from calcined clay with sulfuric acid and to remove the silica residue by filtration. The filtrate from leaching is then purified by one of the following methods: Removal of iron in electrolytic cells using mercury cathodes; removal of iron by chemical precipitation through the formation of an iron-manganese complex; or precipitation of the alumina from an aqueous-ethanol solution, leaving iron in solution. Both the operating and capital costs for these processes are higher than those for the Bayer process. For plants producing 1,000 tons of alumina a day, the estimated operating costs per ton of alumina are: Electrolytic iron-removal process........... $81.35 Chemical iron-.removal process............... 82.08 Ethanol purification process............... 86.17 The investment in mercury is not part of the operating cost since this cost is considered equivalent to catalyst cost which would not be shown. The estimated capital investments for 1,000- ton-per- day plants on a Marshall and Stevens chemical equipment index basis of 240.0 are: Electrolytic iron-removal process...... $116,775,500 Chemical iron-removal process............... 99,418,900 Ethanol purification process................ 94,210,200 These investments include the working capital but not the land for each of the processes. Included in the electrolytic iron-removal process is an investment of $4,859,400 for mercury. Investigation of other methods for removing iron, of the decomposition operation, and of the leaching operation could reduce these costs.

Jan 1, 1963

By L. W. Higley

Procedures were developed for reclaiming high-temperature alloy scrap in the form of standard cobalt-base alloy. A cobalt-base, multicomponent, high- temperature alloy, such as S-816, can be successfully melted from scrap, and its specified chemical composition can be retained. Induction furnace melting in a controlled atmosphere was used to obtain ingots that were subsequently fabricated by impact forging to provide specimens for heat treatment and testing. Of the numerous metal elements in S-816 alloy, only small quantities of molybdenum were lost in the melting operation as finally developed. Test results indicated that the scrap material, after melting, forging, and heat treatment, was equivalent in room-temperature tensile strength, in high-temperature creep and stress-rupture properties, and in chemical composition to commercially produced S-816 alloy made from virgin material.

Jan 1, 1963

By Robert M. Doerr

Isothermal oxidation kinetics were determined for three grades of Ni and for 99.99 percent Co in Ca and in air from 800° to 1,200° C. The oxidation reactions followed approximately the parabolic rate law. For 99.99 percent Ni in 02 at 1,200° C, the parabolic rate constant was 2 mg cm-2 hr-2. For 99.99 percent Co in dry air, the constant ranged from 1 mg cm-2 hr-2 at 800° C to 18 mg cm-2 hr-2 at 1,200° C; for 99.99 percent Co in 02 the constant was 2 mg cm -2 hr-2 at 800° C and 10 mg cm-2 hr-2 at 1,000° C. For 99 percent Ni in dry air, the constant ranged from 0.2 mg cm 2 hr-2 at 800° C to 3 mg cm 2 hr-2 at 1,200° C. By X-ray diffraction measurements, only divalent oxides were found in the scales. Volumetric and quartz-helix gravimetric methods were used. The volumetric apparatus was designed to permit less stringent restrictions on the corrosion specimens; it was also more convenient to use. The agreement between the weight gain determined gravimetrically and the 02 consumption determined volumetrically was not good enough for direct comparison of gravimetric and volumetric results.

Jan 1, 1963

By S. J. Hussey

This Bureau of Mines report presents a discussion of the statistical approach to the design of experiments and indicates, by concrete examples, how and when statistical design is essential for obtaining meaningful data with a minimum number of experiments. Every research worker plans his work and designs his experiments in a manner which, based on his experience, seems most rational to him and which he believes will produce the information essential for resolving the problems. If the results obtained are open to multiple interpretations and are not of definitive reliability, the failure of the experiment design to produce the desired information may be due to several causes, These include faulty judgment in selecting the factors to be tested, ignoring the importance of interaction between factors, failure to adequately control test variables, and assuming that meaningful trends in the results can be identified without comparison to a predetermined experimental error. An experienced researcher probably will deduce from the nature of the data obtained that the design is faulty and may by trial and error ultimately arrive at the correct solution to the problem. Statistical methods of planning research minimize the chance of faulty experiment design, embody concrete procedures for testing the validity and reliability of the test data and for indicating mandatory changes in the design, and enable the researcher to obtain meaningful information with the least number of tests.

Jan 1, 1963

By Hillary W. St. Clair

This is an initial progress report on a theoretical study of the iron blast furnace. The purpose is to develop a thermochemical model that will simulate mathematically a real furnace. An analysis is made of the behavior of a hypothetical furnace, operating adiabatically on pure ferric oxide, graphitic carbon, and dry air. Relationships are derived for carbon consumption, gas composition, and reaction temperature for a furnace operating at steady state, and in which the smelting rate is slow enough for the gas and solids to approach thermal and chemical equilibrium. Analyses are made for two contrasting types of gas flow: (1) Nonturbulent flow, having no vertical mixing; and (2) turbulent flow, having perfect mixing during each step of the process. It is shown that some vertical mixing is essential for complete reduction of ferrous oxide and that the extent of vertical mixing is one of the critical factors in the behavior of the furnace, This analysis indicates that the thermochemical relationships in a blast furnace can probably be best simulated by a finite series of gas-solid reactors through which gas and solids flow countercurrently and in which the temperature and composition of gases and solids are assumed to be uniform. Any degree of vertical mixing can be approximated by the proper choice of the number of reactors.

Jan 1, 1963

By R. D. Berkenkotter

Titanium and iron assays obtained from cores of three diamond-drill holes from the Cebolla Creek titaniferous iron deposit were evaluated by statistical techniques to obtain information concerning the mineral characteristics of the deposit and to make a preliminary determination of sampling requirements for the deposit. A range of sample volumes representing sizes from one-sixth NX core 2 feet in length through full NX core 10 feet in length were prepared by split-ting the NX core with a diamond saw and combining pulverized splits. Results indicate the existence of at least two statistical populations having different average grades of iron and titanium. Each has an assay frequency distribution that is not normal but is highly peaked (leptokurtic). Frequency distributions of assays from specific altered zones of the drill holes appear to be normally distributed. Increasing the cross-sectional area of the sample unit to change assay volume did not result in significant changes either in assay values or the resulting frequency distributions of assays. However, when assay volume is increased by increasing sample interval, the shape of the assay frequency distribution changes, and a normal distribution of assays is approached. Positive correlation exists between titanium and iron assays. The degree of correlation, although very significant, is too low to use for prediction purposes. Although the amount of data from three widely separated drill holes is not sufficient to establish an average grade for this large deposit, it indicates a range of expected assay values and provides some information for determining requirements of future sampling for evaluation purposes.

Jan 1, 1963

By Robert C. Johnson

Water-swelling fluormicas and fluormontmorillonoids were synthesized and described. They were found predominantly in the system Mg0 - MgF2 - Si02, with varying amounts of Li+ or Na+, Some water-swelling synthetic minerals were prepared with only strontium or lead in the twelvefold position, Water- swelling compounds containing either K+ or Ba+ were not found. The synthesized water-swelling products were shown by electron microscopy, X-ray diffraction, and chemical analysis to be typically two-silica-layer, platy compounds, Except for the complete substitution of fluoride for hydroxyl, the compounds were similar to the natural micas or montmorillonites. The stability and properties of the compounds were further established by electrodialysis, by differential thermal analysis, and by anion and cation exchange methods. Fluoride in the interlayer positions was nonexchangeable except through decomposition of the product. Exchanging the sodium or the lithium of the synthetic fluormicas or fluormontmorillonoids with potassium ion, particularly those derived from KOH provided a product from which a completely inorganic paper of extraordinary flexibility was fabricated. This paper had excellent mechanical and electrical properties except for water adsorption and a high power factor. Neither of these problems was solved during this investigation. However, papers with dielectric strength from 1,500 to 5,000 volts per mil were readily fabricated, particularly from the lithium containing fluormicas or fluormontmorill onoids.

Jan 1, 1963

By Carl Rampacek

The conversion of dilute sulfur dioxide gas into sulfuric acid by autoxidation in mechanical cells of the flotation or gas absorber types was investigated. Bench-scale tests were made with sulfur dioxide-air mixtures to determine the influence of several factors on autoxidation of sulfur dioxide, using iron and manganese salts as catalysts. The laboratory investigation subsequently was extended to pilot plant acid-making tests, using a modified turbomixer gas absorber having a 12-inch-diameter impeller. Sulfur dioxide gas was produced for the tests by roasting pyrite in a fluidized-bed reactor. In the better tests, made with manganese sulfate as the catalyst, the equivalent of 1,000 pounds of anhydrous H2SO4 was produced per 24 hours as a 10-percent solution; power consumption was 0.2 kwhr per pound of equivalent anhydrous acid synthesized. The batch and continuous tests demonstrated the technical feasibility of producing dilute sulfuric acid by autoxidation of sulfur dioxide gas in mechanical cells when converting pyrite roaster or waste smelter gases.

Jan 1, 1963

By H. C. Carpenter

Lignite tar made by low-temperature Lurgi-Spülgas carbonization of North Dakota lignite was hydrocracked in single-pass, fixed-bed operation at 3,000 pounds per square inch over cobalt molybdate catalyst to produce gasoline, and at the same pressure, over zinc chromite catalyst to crack the high-boiling phenolic compounds to low-boiling phenolics. Hydrocracking the raw tar or the creosote-oil fraction from the tar over cobalt molybdate removed most of the oxygen, sulfur, and nitrogen in the forms of water, hydrogen sulfide, and ammonia, and produced gasoline yields from 31 to 59 volume-percent. The gasoline fractions contained from 0.1 to 7.7 volume-percent tar acids, but tar-base percentages were too low for reliable measurement; removal of these materials with dilute caustic and acid solutions also reduced the sulfur and nitrogen to 0.01 weight-percent or less.

Jan 1, 1963

By Robert S. Warfield

The Bureau of Mines investigated a geologically favorable area in the Beluga River coalfield in south-central Alaska to determine if there were sufficient minable coal reserves for onsite utilization to meet the projected thermal power needs of the Anchorage area. A program of diamond core drilling and sampling delimited a shallow saucerlike structure estimated to contain indicated reserves of more than 20 million tons of subbitminous coal in a seam over 50 feet thick that could be mined by opencut methods. A large sample of fresh coal was taken across the full width of the seam for washability studies. This sample, with 4 volume-percent of clay partings eliminated, contained 24.4 percent moisture and 16.8 percent ash and had a calorific value of 7,162 Btu per pound. The studies indicated that mechanical cleaning would be of doubtful value.

Jan 1, 1963

By A. W. Deurbrouck

Five preparation plants employing concentrating tables for washing 3/8- inch to 0 raw coal were tested. Two of the plants were washing Pittsburgh-bed coal, two were washing Pocahontas-bed coal, and one was washing Elkhorn-bed coal. Samples of the feed to and products from the tables were collected over the normal operating shift. These samples were analyzed, and complete performance criteria were determined. The results of this study show that the concentrating table is an effective fine-coal cleaner even when operating at 40 percent over the recommended 10 tons per hour feed rate. Recovery efficiencies for the 3/8-inch to 200- mesh size fraction ranged from 97.3 to 99.3 percent; maximum yield errors and ash errors were 2.4 and 1.0 percent, respectively; probable errors ranged from 0.074 to 0.115 specific gravity units and error areas from 52 to 73 square centimeters. The amount of minus 200-mesh material in the water used for feed water and dressing water to the table had no significant effect on the sharpness of the separation achieved but did have an effect on the ash content of the clean- coal product. In the one plant studied, the substitution of fresh water for plant circulating water reduced the ash content of the clean coal by about 6 percentage points.

Jan 1, 1963

By A. R. Taylor

The Bureau of Mines has determined the heat capacity and enthalpy values for beryllium sulfate (BeS04) throughout the temperature ranges of 10° to 300° K in an adiabatic calorimeter, and 350° to 860° K in a Bunsen-type ice calorimter, The heat-capacity data were extrapolated to 0° K, and smooth values of heat capacity, entropy, enthalpy function, and free-energy function were reported at 10-degree intervals between 0° and 300° K. The method of least squares was used to fit the enthalpy data with an algebraic equation. The equation also was used to calculate values of heat capacity, entropy, enthalpy function, and free-energy function at 50-degree intervals between 300° and 900° K.

Jan 1, 1963

By L. B. Pankratz

The heat content above 298.15° K was determined for crystalline vanadium silicide of the composition V3Si. This substance has a minor thermal anomaly (probably a heat-capacity maximum) near 560° K; the extra heat absorption is about 50 cal/mole. The original heat-content data are presented, and smooth values of the heat content and the corresponding entropy increments are listed for the temperature range from 298.15° to 1,500° K. Equations representing the heat- content values and derived heat capacities are also included.

Jan 1, 1963