Search Documents

By Shiou-Chuan Sun

There are several postulations for the mechanism of slime-coating. Incel proposed the electrostatic hypothesis, del Giudice2 suggested the chemical theory; Bankoff3 reported that slime-coating is inhibited whenever the surfaces of the slime particles are ionized sufficiently for Brownian movement to occur, and is facilitated when they are not. These postulations were examined by experiments involving the Burton tube, the cataphoretic cell, and slime-coating. The techniques involved will be described elsewhere in this paper. The electrostatic hypothesis, supported by migration experiments in a Burton tube, was abandoned by del Giudice.2 Experiments of mineral migration in a Burton tube. after the idea of Ince with modification, indicated that this apparatus was inefficient for accurate observation. A cataphoretic cell was constructed. The experimental data, as shown in this paper, enable the correlation and advancing of this hypothesis. Del Giudice's chemical hypothesis considered that slime particles are attached to the galena surfaces by a cement formed as the result of a metathetical reaction between the sulphide and the slime particles. Applying this hypothesis, the calcite slime particles were cemented to the galena surface by the formation of a cementing lead carbonate. In a similar way, lead silicate was considered as the cement of quartz slime coating on galena. Experiments verifying this hypothesis were carried out according to the following procedure: Fresh mineral particles selected from the central part of the broken lumps and examined first under a microscope for ascertaining a clean surface were placed in a small beaker (1.5-in. dia. by 2.5 in. deep) half filled with one kind of slime pulp. The dilution of all slime pulp was 4:I and the particle size of slime was minus 400-mesh. After agitation for 5 min. the specimen was removed by a sharp-end tong and washed gently three times in three separate distilled water baths. It was dried before a small electrical fan and examined immediately under a microscope. A photograph was made of the observation. With reference to the published solubilities,' several of these experiments, which contradict the chemical hypothesis, are listed as following: I. The coating of fluorite slime on the galena surface was light but the coating of galena slime on the fluorite surface was heavy. In a similar way, the coating of quartz slime on the galena was fairly light, but the coating of galena slime on the quartz particle was heavy. According to the chemical hypothesis, using the same constituents and regardless of which is the slime, the density of the coating should be expected to be the same. This, however, as shown above, is not true. Moreover, fluorite (CaF2), having a solubility of 16 mg.4a per liter of water, should not be

Jan 1, 1943

By Arthur F. Taggart, A. W. Thomas, T.C. Fitt

It was noticed early by operators that high recoveries and flocculation of the sulphide minerals were closely correlated in agitation-froth flotation. Later, this readily visible flocculation was found almost invariably to be due to the binding together of the oiled sulphides by minute air bubbles, and is now known to be characteristic of pulps in which at least one of the mineral species is coated with an insoluble collecting oil. In 1930 it was reported1 that in differential flotation a correlation exists between Brownian movement of particles in the pulp and their nonfloatability, and that Brownian movement and dispersion are likewise correlated, whence it would appear to follow that dispersion of a mineral particle in a flotation pulp denotes a particle condition unfavorable to flotation. In 1934 it was postulated2 that Brownian movement, dispersion and nonfloatability all flow from the condition of the surface of the dispersed particles. It was also postulated at that time that the surfaces of dispersed particles are ionized, the ionized compound being a product of reaction between a component of the particle and a component of the pulp, of a mass solubility of the general order of a fraction of a milligram to I 5 to 20 mg. per liter. It was then further asserted, and has since been confirmed by Bankoff,3 that sulphides so conditioned with soluble collectors as to be floatable flocculate without the intervention of air bubbles as binding media. The present paper comprises an inquiry into the nature of the bound and counter ions in zinc sulphide dispersed by sodium arsenate. A part of its findings was mentioned in the 1934 paper above cited.= The tests show that: I. When powdered granular zinc sulphide is immersed in aqueous solutions of sodium arsenate, arsenate ion is abstracted from solution and anchored to the solid surface, while sulphate and lower sulphoxy ions are concurrently displaced therefrom in substantially stoichiometric quantities. 2. The ion counterbalancing the anchored arsenate ion is zinc. 3. Dispersion and Brownian movement of zinc sulphide are concurrently maxima at an arsenate-ion concentration lying between 0.01 and 0.001-molar sodium arsenate. 4. Any condition or reaction that tends. to remove the zinc arsenate coating from the zinc sulphide particles decreases dispersion and Brownian movement concurrently. Experimental Three samples of zinc sulphide were used. Lot I was C.P. zinc sulphide, washed well be~ond absence of chloride, as shown by test with silver ion. Lot 2 was prepared by precipitation from hydrochloric acid solution of arsenic-free zinc, and was washed as lot I. Lot 3 was prepared by heating a portion of lot 2 in absence of air at 1200° to r300°C. for one hour."

Jan 1, 1943

By Arthur F. Taggart, Nathaniel Arbiter

The fact that the floatability of minerals with fatty-acid collectors changes as the pH of a pulp varies was utilized in the early days of flotation, when sulphuric acid was used with oleic acid to float sphalerite and galena. from gangues which themselves float with fatty acids in alkaline pulps. Gaudin1 showed that this difference in floatability is a regular function of the extent of the change in pH. The present paper offers an explanation of the phenomenon, and on the basis thereof sets up recommendations for control of soap-flotation operations. Experimental Parallelism between floatability of a mineral in a given environment and its contact angle in that environment is so well established,2 and the contact angle is so readily and rapidly measurable, that contact angles have been used throughout the experimental work to indicate relative floatability. Experiments comprised contact-angle measurements on a number of representative minerals and metals, and a variety of subsidiary tests designed to check and to amplify the contact-angle indications. Contact-angle measurements were made by the method described in detail by del Giudice.3 Sequence of procedure for determination of the points of Fig. I was: I. Polish specimen. Most specimens yielded zero contact angle after polishing; a few showed a slight "cling." Cling is characteristic of specimens that pit on polishing; it is due to a residue of organic contamination, originally on the particle surface, which persists in the pits. The cling did not change total contact angle after conditioning, and is not, therefore, additive in the upper part of the curves of Fig. I. It may have broadened the bases of the curves for the affected minerals slightly. 2. The specimen was next conditioned in soap solution for 5 minutes. 3. The conditioned particle was transferred, under conditions that precluded surface drying, to an aqueous solution of the pH desired and contact angle determined. 4. The particle was thereupon transferred, again without surface drying, to water. The confirmatory tests will be described at the places pertinent in the argument. Discussion of Results Gaudin and Vincent4 have shown that contact angles with heptoic acid as a collector decrease for a given specimen when the concentration of collector present falls below a certain value. A similar reduction in angle occurs with collector-conditioned particles that are visibly slime-coated, and varies roughly with the extent of coating. Since in the last case the reduction in angle is certainly due to reduction in the proportion of total conditioned mineral surface presented to the bubble, the authors believe that the same condition prevails in the first case cited, and that similar reduction in angle and floatability will occur whenever starvation quantities of collector are present, whether by design or by reason of excessive consumption.

Jan 1, 1943

By Dwight L. Baker, Alexander Knoll

The work herein reported shows that galena in certain organic solutions of fatty acids becomes coated with lead soaps, and that this coating is not only highly water-repellent but is also repellent to certain organic liquids, particularly those of high surface tension; i.e., nitrobenzene, diphenylmethane. Wetting Properties of Galena in Nonaqueous Fatty Acid Solutions When a "captive bubble"I is withdrawn from a polished, plane galena surface under a nitrobenzene solution of stearic acid, the air bubble shows a decided tendency to "cling" to the mineral surface. The contact angle, therefore, was greater than zero, although it was not large enough to be measured quantitatively by the captive-bubble method." When pure nitrobenzene was substituted for the nitrobenzene-stearic acid solution, no clinging of the bubble was observed; viz., contact angle = zero. Although the contact angle developed in air-galena-nitrobenzene-fatty acid solution was small, the organic solvent repellency of the galena surface was great enough to enable good recovery in a laboratory flotation cell2 In Table I are given the results of flotation experiments using several different fatty acids dissolved in nitrobenzene.

Jan 1, 1943

By M. D. Hassiallis

The claim is made in a number of patents1'2'3'4 that some compounds of the class known as aryl sulphides have collector properties. One of these patents generalizes the claim to include all aryl sulphides, whose generic formula is R-S2-R' (where x = 1,2, . . . ), with the restriction that if the aromatic radicals R, R' contain substituent groups these groups may not contain nitrogen or oxygen. Moses, Hess, and Perkins cite as evidence for their claims the results of flotation tests on Anaconda and Miami copper ores using diphenyl sulphide, ditolyl disulphide and dinapthyl disul-phide as collectors. The use of alkyl sulphides as collectors has also been reported5 with the conclusion that the alkyl sulphides are better collectors than the corresponding aryl sulphides. The experimental work cited appears to confirm this contention. In order to test these claims with respect to diphenyl sulphide, a sample, purchased from the Eastman Kodak Co., was tested against Anaconda ore. The procedure followed. was that of Moses, Hess, and Perkins. A sample of ore weighing 500 • grams was placed in a laboratory ball mill. To this was added 220 grams of water, 0.06 grams of commercial diphenyl sulphide (1/4 lb. per ton of ore), and 0.8 grams of lime (3 lb. per ton of ore). The material was ground until 93 per cent passed a 65-mesh screen. This pulp was diluted to 16 per cent solids. A sample of the diluted pulp was placed in a Minerals Separation type of laboratory flotation cell. After a a-min. agitation period, pine oil was added and the froth collected for a period of 15 min. This test recovered 68 per cent of the copper, with a ratio of concentration of 6.8:I. The tailing assay was 0.3 per cent Cu. Moses, Hess and Perkins attained a recovery of 75 per cent, with a ratio concentration of I2:I. Although these results do not compare too well with those of Moses, Hess and Perkins, subsequent tests showed an improvement, as judged by vanning of the tails. It is concluded, therefore, that the patent results are confirmed. Contact-angle Tests In order to study the response of the various minerals, found in Anaconda ore, to commercial diphenyl sulphide, some contact-angle tests were made. Polished particles of bornite, chalcodte, chalco-pyrite, pyrite and quartz were conditioned for 15 min. in a suspension of diphenyl sulphide in water. The concentration of this suspension was 0.284 grams per liter, this being the concentration of this reagent in the ball mill. The conditioned particles were rinsed with water (not rubbed with filter paper or boiled linen) and then tested against an air bubble in the bubble machine, with the following results: born-

Jan 1, 1943

By Arthur F. Taggart, Herbert H. Kellogg, Nathaniel Arbiter

The experimental work herein described is presented in support of the following broad hypothesis: Conditioning of metals and metallic sulphides by amine collectors involves metathetic reaction at the solid surface between the collector and an oxidized metallic compound, to produce a coating comprising an amine-bearing compound of low solubility in the system. The type of reaction is different in acid and in basic solutions; both types may occur simultaneously in near-neutral pulps. Experimental The experiments comprise contact-angle tests with a variety of metals and metallic sulphides and extraction teats on ground chalcocite. Contact-angle tests were made principally on metals. Covellite paralleled metallic copper in amine solution at pH 8.8 (Table 3) and in alkaline water, but maintained an angle of about 70° in Io-3 HCl solution, where metallic copper lost its angle. The tests with anionic collectors (Table I) indicate that the behavior of metals in general is parallel to that of their minerals in anionic solutions. It is believed that this is generally true with cationic collectors also, but the covellite test is a warning against a broad assumption to this effect. All test specimens were mounted in Transoptic.* All tests in conditioning solution were preceded by a test in distilled water; cleaning and polishing were continued in every case until the angle in water was zero. Subsequent transfers into and out of solutions and test cells were made in such a way as to prevent exposure of the surface to air.1 The amine used was primary laurylamine, Armour and Co., AM-1120. It was used in the form of the hydrochloride, obtained by precipitation with gaseous hydrochloric acid from an ether solution of the amine, and purified by two recrystallizations from ether-alcoho1.2 Inorganic and other test reagents were standard laboratory grades. Temperatures were not controlled; the probable range of test-solution temperatures is 650 to 80°F.

Jan 1, 1943

By S. F. Ravitz

In his excellent book on the Principles of Flotation, Warkl makes the following significant statement concerning the theory of flotation: Two questions of first-rate importance must be considered . . . First, it must be decided whether the sulfide minerals require a collector in order that they may float, or whether as has been claimed, they possess an inherent float-ability and need no collector. Second, it must be decided whether the xanthate type of collector can react with or be adsorbed by the unchanged sulfide minerals themselves, or whether, as many claim, reaction with the collector is dependent upon initial oxidation of the mineral surface. In the theory of flotation there are no two questions of greater importance than these. Ravitz and Porter2 presented evidence that galena can be floated without a collector, and even without any reagent, provided oxidation products are thoroughly removed from its surface. They found also that thoroughly cleaned galena floated faster when xanthate was used than when a frother alone or no reagent was used, and that, without a collector, the recovery decreased gradually upon exposure of the mineral to air. Their work, however, has met with considerable criticism.l(127).3-6 This paper is presented to answer the objections that have been raised, and to present additional data that confirin their results. The apparatus used to provide oxygen-free conditions for flotation has been described in detail by Ravitz and Porter. An atmosphere of purified nitrogen was maintained within the apparatus; all water and solutions used were freed of dissolved air, and could be transferred where desired by the use of nitrogen Pressure; nitrogen was used as the gaseous medium for flotation. The flotation cell was essentially a glass cylinder 15 cm. long and 4.6 cm. in diameter. Near the bottom was sealed a fritted glass filter, which broke up the gas into small bubbles, and 6 cm. above this filter was a side arm through which the froth overflowed. The pulp was agitated by a glass impeller, which entered the cell through a mercury seal. For each test 17 grams of galena and 75 ml. of water were placed in the cell, after which the speed of the impeller and the rate of gas flow were adjusted. Water was added gradually during the test to keep the froth level at the proper height. The mineral was prepared and cleaned in the following manner. Exceptionally Pure galena was crushed by hand on a clean bucking board, screened to —I50+ 200 mesh, and deslimed on a 200-mesh sieve with a fine spray of water. It was then boiled on a hot plate for a number of hours with several changes of the desired cleaning solution, after which both the solution and the mineral were placed in the "mineral flask" of the apparatus. The cleaning solution was drained off, and the mineral was boiled with oxygen-free water

Jan 1, 1943

By C.H. Lorig, M.C. Udy

Through the years much interest has been centered in attempting to develop a direct method of iron-ore reduction, to replace or supplement the present indirect blast-furnace process. It would not be difficult to produce a list of more than 500 patents dealing with the direct production of iron and steel, but the few of this type of process that have proved successful do not produce in competition with the blast furnace; they produce special products. A survey of the literature reveals that the equilibrium conditions for the reduction reactions (iron oxides with carbon monoxide and with hydrogen) have been fairly well established but that data on the rate of reaction are discordant, although the genera1 effects of the variables are known. This dissertation deals with the rate of reduction of a magnetite concentrate with hydrogen. The following variables have been studied: bed depth, gas velocity, particle size and temperature. The behavior of a single particle was also studied. Literature Two general factors should be considered in a study of reduction of iron oxide: (I) equilibrium relations, and (2) rate of reaction. The first factor, equilibrium, has received the attention of many investigators and the equilibria Fe304-FeO-H2-H2O, Fe3O4-FeO-CO-CO2, FeO-Fe-H2-H2O, FeO-Fe-CO-CO2 are fairly well established. Of the second factor, the rate of reaction, much less is known. Data are comparatively scarce and there is very little concordance. The discordance probably results from the large number of variables influencing the reactions. However, the general effect of the variables is known. Among these variables are temperature, gas velocity, particle size, porosity, gas composition (distance from equilibrium), presence of impurities, greater or less prevalence of side reactions, and even design of the apparatus itself. Wetherill and Furnasl published an excellent paper in 1934. They point out that any one or more of five processes (the chemical reaction, diffusion of the reactant through the gas film, diffusion of the reactant through the solid phase, diffusion of the reaction product through the solid phase, or diffusion of the reaction product through the gas film) may be the slow process, and thus the rate-controlling process. It is regretted that the work of these authors was limited to experiments at One temperature, and that the range of gas compositions was limited. Langmuir2 has shown that heterogeneous reactions take place at phase boundaries.

Jan 1, 1943

By Frank G. Norris, William C. Marshall

The term "total operation" is meant to include problems that cannot be answered from the standpoint of either the blast furnace or the open hearth separately but must be studied by considering the interrelations of these two processes. There are other ways of expressing this idea, such as reference to the blast furnace and the open hearth as an integrated process or by emphasizing the interrelation and interaction of the two processes. Much progress has been made in metallurgy and other fields by breaking complicated processes and reactions into simpler parts and studying each of these intensively. This intensive study frequently is done under the controlled conditions of the laboratory experiment. The entire process is understood better by this specialized study of each of its parts. This procedure is so satisfactory that an understanding of unit operations is a fundamental requirement of chemical engineering. Much discussion of the open-hearth process is in terms of one heat or one furnace. By this is meant that the conclusions apply to one furnace, wherever it may be. Furnaces seldom occur singly; they are located in groups and these groups are located in steel plants having characteristic blast-furnace equipment. Problems that are affected by the number of furnaces operating, by the number of blast furnaces in the plant and by the fact that the open- hearth shop is dependent on the bla furnace capacity rather than on its o needs for the amount of iron available among those which, in contrast to "u operations," we have chosen to call "to operations." Many of these problems, such as cha ing delays and the general mechani problem of material handling, are familiar that merely to define " total op ations" brings them to mind. These pr lems are of such importance that alrea they have received careful thought 2 study and probably will merit furt study. As will be shown later, these mechan problems are so intimately related metallurgical problems of total operati that the effect of certain changes cannot ascribed with certainty to either a mech ical or a metallurgical cause entirely. Although the problems coming under general term of "railroading" are inclu in the problems of total operations, ' others are selected for the discussion of paper, One of these is strictly a problem composition; the other is strictly a prob of tonnage production. The problem of phosphorus is pres whenever basic open-hearth slag is used the blast-furnace burden. This practice not secured universal acceptance chi because of difference in preference for h manganese or low-manganese pig iron hjgh-manganese (2.00 per cent) iron is to made, open-hearth slag as a source manganese is usually cheaper than m

Jan 1, 1943

By C.A. Zapfee, C. E. Sims

An investigation of the behavior of inclusions in steel several years ago1 led to the conclusion that some of the commonly occurring inclusions in steel have appreciable solubilities, particularly in liquid steel. In the belief that further study and clarification of this matter was warranted, a laboratory study of oxides in liquid steel was started. These oxides are generally the result of deoxidation treatment, and the basis of an understanding of their behavior must be, therefore, the behavior of the deoxidizer in liquid iron. The Fe-Si-O system was selected for the first study, and the apparatus used and experiments conducted are described herein. Apparatus The apparatus is shown in Fig. r and the system is shown by diagram in Fig. 2. Furnace, A lmosphere Commercial hydrogen (1) passes through an orifice (2) into a manometric system that measures the tankside pressure (3) and the back pressure (4) in the system. A water trap (5) guards against excess tankside pressure. A by-pass with stopcocks (6 and 7) is provided, so that the hydrogen can be routed directly from the tank to the furnace if desired. When stopcock 7 is closed and stopcock 6 is open, the gas proceeds through the conditioning train, of which the first unit is a furnace (8) that contains platinized asbestos at 350°C. The temperature is measured by a thermocouple and potentiometer (9) and is controlled by a slidewire resistance (10). In this furnace fres oxygen is converted to water vapor. The exiting gas passes over a water bath (11) kept slightly below boiling temperature by an auxiliary heating unit (12). This unit also heats a surrounding copper coil (13) containing water, and the heated water rises through a collar (14)

Jan 1, 1943

By C.R. Taylor, John Chipman

The study of chemical reactions of liquid steel and basic open-hearth slag involves a complex slag system of at least eight important components, and often a number of others. In initiating an experimental program on the equilibrium relationship of slag and metal, it became evident that but little progress could be expected in the interpretation of the complex slags until an understanding had been gained of the simpler liquid systems composed of a limited number of slag constituents. The simplest system that in any way resembles open-hearth slag must contain lime, silica and iron oxide. Attempts to produce in an induction furnace slags containing only these constituents have met with serious limitations because of the solubility of refractories in the slag. Thus in acid-lined furnaces the slag is always saturated with silica and consequently contains between 48 and 63 per cent of this component. In basic-lined furnaces an undesired additional component, magnesia, is introduced to the extent of some 3 to 15 per cent, depending upon the acidity, sometimes with disastrous results to the furnace. Thus in a magnesia-lined furnace, Fetters and Chipman1 found it impossible to prepare slags in which the ratio (CaO + MgO)/SiO2 was less than that corresponding to the metasilicate (approximately 0.8:I). Between this composition and that of the true acid slags lies a field of composition about which very little is known and which cannot be explored by any ordinary means. This circumstance, as well as the desire to study the more basic slags in the absence of magnesia, calls for the development of a method by which slags may be brought into equilibrium with molten metal without coming in contact with refractory. Such a technic has been provided by the rotating induction furnace developed by Barrett, Holbrook and Wood.2 Three modifications of the rotating furnace were described by these authors. The first was essentially a rotating refractory cylinder containing the magnesia crucible, which was itself heated by a graphite crucible. In the second, only the magnesia crucible rotated, the cylindrical graphite heater being stationary. In the third, the coil and crucible revolved, and the metal was heated directly by induction. Experimental Method Furnace The furnace used in this investigation differed from previous designs and may be Sci- a fourth modification of the original rotating furnace. Its essential fea tures are shown in Fig. I' The Outer case A is a fire-clay flue liner for control of the atmosphere. The top B and base 6 are alberene stone. The coil supports D are of ""Formica," a material formed from phe-

Jan 1, 1943

By Gerhard Derge

The extension of metallurgical control of steelmaking processes has always made it desirable to have some quick method for determining the oxygen content of molten steel. To meet the practical demands of the steel industry, this method must also be simple enough so that the sampling can be done by the regular furnace crew. This requirement was kept in mind continually in the work to be described. An oxygen value is required for proper control, whether it be of a rimming steel or of a fully killed steel. However, two factors have discouraged the development of such direct methods: i. The reaction C + FeO Fe + CO is rapid enough to make the correct sampling Of the bath a difficult problem. The most generally accepted means of circumventing this difficulty have been the aluminum-killed bomb test of McCutcheon and Rautiol and the aluminum-killed spoon test of Herty.2 The theory of these tests is that all of the oxygen in the iron is converted to Al2O3, which can then be determined by chemical methods. Even when simplified so that the analysis is made by a measurement of turbidity, this method is too slow to be used in routine control. Furthermore, the results obtained have never been entirely satisfactory, and some of the reasons for this will be discussed later in this paper. z. The ordinary chemical methods of analysis for oxygen in steel are not reliable. The only analysis that seems to possess sufficient dependability to be of value is vacuum fusion,3 and the apparatus and technique developed for this have been too complex and slow for routine work. Indirect Methods Indirect methods of estimating the oxygen have been resorted to in mill practice. Generally these depend upon the chemical equilibria involved. The oxygen in the bath is related to other components in the slag or metal that can be analyzed the readily. These methods have many difficulties in common. Steelmaking processes involve a complicated set of interrelated reactions, all of which tend to approach equilibrium at varying rates. The degree to which any particular equilibrium is approached is not well established. Moreover, a proper use of theoretical equilibrium constants requires a more extensive knowledge of the activities of the con- than we now have. This is well illustrated by the irregularities observed when trying to determine the oxygen in the metal by its ratio to the iron oxide in the slag. Equiliblium relations may be used mole readily if they are evaluated by plant experience, but the extent of approach to equilibrium remains an undetermined variable and lack of an entirely satisfactory mean, of determining oxygen for the calibration is felt keenly, For example, the statistical analysis of the relation between carbon and oxygen4.6 illustrates that even iia a single plant wide variations exist under llormal conditions of operation. Never. theless, this is one of the best methods now

Jan 1, 1943

By Margaret Dienes, Karl L. Fetters

It is recognized that ingot structure is important in determining the quality of finished steel. Such elements of ingot structure as the size, shape and distribution of primary crystals; the size, shape and distribution of shrinkage cavities; the degree and kind of segregation (both macro and micro); the distribution of nonmetallic impurities and gas cavities are each the result of solidification phenomena. 'These structural features vary with variations in ingot practice. Their control requires a knowledge of their correlation with the variables of ingot practice. The structures of full-sized commercial steel ingots have been studied extensively by the Steel Ingots Committee of the Iron and Steel Institute.l This work has contributed much to our knowledge of ingot structure and the factors that control it. These investigations of necessity have been limited in scope because of their costliness and the great amount of time required. Studies of the structure of small steel ingots have not been very satisfactory because of the rapid rate of freezing of the ingots. Their structures are thus difficult to resolve and to interpret. Medium sizes of steel ingots in the ranges of 150 to 500 lb. do offer opportunity for study, but their cost also becomes greater and their use less expeditious. Experimental models for the study of steel-ingot structure have been made of nonferrous metals. Gathmann2 used lead-tin alloys and reported that they were satisfactory for determination of the location of the major shrinkage cavities and the Liplanes of weaknessv resulting from the pattern of crystallization. Certain features of mold design have been illustrated by these results. Bezdenezhnihk3 has used copper and aluminum ingots similarly and has shown the effect of mold-wall thickness on the depth of columnar grains. These studies of nonferrous models have been useful, but their results have been limited because the analogy with steel ingots obviously is not a close one. Model wax ingots have also been useful in pointing out important features of mold design. Gathmann4 watched the solidification of wax ingots, through a glass-walled mold, and has reported the effects of mold type on the location of the shrinkage cavities. Segregation was simulated by using a mixture of paraffin and stearine and adding a red dye that on freezing segregated preferentially to the stearine. Brearley5 also used wax models to determine the effects of some of the principal features of mold design. Carlsson and Hultgren6 dsed stearic acid to show that the columnar crystals of ingots tend to slope upward progressively from the mold wall toward the center of the ingot. As with the nonferrous ingots, the wax ingots have

Jan 1, 1943

By T. J. Woods

This paper deals with the origin, definition and prevention of scabs on semifinished rolled-steel product. Mold coatings, which are considered essential in scab prevention, were found to be effective only in relation to the ingot surface of semikilled and killed steels and not necessary for the rolled product. That is to say, the scabs on ingots caused by having been cast in uncoated molds are eliminated in the soaking and rolling operations, therefore are not present on the rolled product. This fact, which is contrary to the average opinion, was established after extensive and thorough experiments had proved the ordinary conception of the origin of scabs to be incorrect. This subject may be of little or no concern to steel plants producing small ingots for the manufacture of small billets and bars, except as it may be of interest in showing why their surface troubles with scabs do not exist on the product of heats with normal pouring practice. But to the producer of large ingots to be rolled into slabs for large plates used in shipbuilding, it is a very important subject. Definition of a Scab The generally accepted opinion is that the individual scabs present on rolled products (as shown by Fig. I) are considered to have been individual splashes on the ingot surface. This opinion was found to be incorrect. Experiments showed that there were no individual scabs on the product that had appeared as individuals on the ingot surface. The single scabs seen are either pieces of a shell that is present near the bottom of all top-cast ingots, or a veneer that may be present on any part of an ingot improperly poured, as when the stream strikes the mold wall. This definition is based on a detailed study of scabs and their characteristics in which the following have been found true: Location.—On the product of heats poured without trouble, scabs, if present, are always on the product that corresponds to the bottom end of the ingot, and usually occur in clusters. This is true whether molds are coated or not. Shape.—There is no regularity as to shape, size or thickness. It is sometimes possible to separate a few pieces and place them together, like pieces of a jigsaw puzzle, indicating that at one time they had been united as one piece. If individual splashes were the origin of scabs, the foregoing would not be true. Instead, the splashes, which occur throughout the pouring of the ingots, would be present uniformly over the entire product and there would be considerable similarity in the shape and thickness of the scabs. Experimental Work Our experimental work was conducted on 200-ton heats of semikilled plate steel tapped into an oval ladle 15 ft. high. All ingots were top-cast direct into 24 by 54 by 82-in. molds, straight-sided, set on flat stools. Two separate studies were made about one year apart, each study extending

Jan 1, 1943

By C.A. Zapffe

Both the foundryman and the theoretical metallurgist are now generally agreed that the anomalous "rising" or "bleeding" of certain ferrous castings of killed metal is primarily attributable to hydrogen. How and why the phenomenon occurs, however, have not been made clear. The following discussion is intended to harmonize foundry experience with metallurgical theory so that the foundryman may possibly improve his ability to prognosticate whether a certain casting may bleed or not. Bleeding, Rising, and Piping The "rising" to be discussed here is not to be confused with the action in rimming ingots as a result of the carbon-oxygen reaction, although a conclusion will be drawn that may apply to the art of controlling the rimming action. Only killed metal is considered, which includes both steels and irons. In Fig. I, three killed, chill-cast ingots are shown: a porous ingot whose evolution of gas during late stages of solidification has ejected metal through the top; an ingot in which moderate gas evolution during solidification has lifted the mushy central metal, i.e., it has "risen"; and, third, a sound ingot with good piping characteristics. The first type is referred to as "bleeding," and generally infers that a solid skin had first formed through which the liquid metal later erupted. There is no fundamental distinc- tion between bleeding and rising in reference to killed metal.* Porosity Hydrogen appears as a cause of bleeding because the gas evolution simultaneously leaves a porosity of a type traceable to hydrogen. Twenty years ago, Melmoth' noted that killed steel that rose unexpectedly in the mold, if returned while liquid to the furnace and heated further under a reducing lime slag, would again rise in the mold; but that if the ingot were allowed to solidify and was then remelted no rising occurred. A critical quantity of the responsible gas is therefore removed by the rising action. Melmoth associated rising with high silicon (0.7 to 0.9 per cent) and concluded that silicon increased gas absorption. Because silicon is a strong deoxidizer, oxygen must, therefore, be eliminated from consideration of the three common gases in steel—oxygen, nitrogen, and hydrogen. Butterworth2 later decided that pinhole porosity was caused by gases dissolved in the metal and that hydrogen was obtained from steam in the mold. Allen's classical work on copper castings,3 soon followed by

Jan 1, 1943

By Frank S. Gardner, Morris Cohen, Dara P. Anita

There is a conspicuous lack of information in the literature on the precise role played by residual quantities of austenite in heat-treated steels. While retained austenite may be expected to have significant effects on such important properties as the impact resistance of tools and carburized parts and the dimensional stability of dies and gauges, attempts to study these problems systematically have suffered from the absence of a reliable technique for ascertaining absolute amounts of austenite. With such a technique, not only could the relation between residual austenite and service behavior be investigated, but the chemical and heat-treatment factors that control the quantity of the austenite could also be established. Thus, the retained austenite content could then be regulated in accordance with the results desired. The presence of retained austenite may be inferred by observing the changes in properties (hardness, specific volume, electrical resistance, etc.) due to its decomposition on subatmospheric cooling1 or on tempering,2 but these indirect measurements are not readily translated into percentages of austenite. Quantitative determinations have been achieved in special cases by means of dilatometric3 and magnetic4,5 measurements; but the former procedure requires that observations be conducted during the hardening and involves difficult calibrations that limit its applicability, while the magnetic procedure is very sensitive to extraneous influences and depends upon assumptions that can only be considered as approximate in an absolute sense. Furthermore, both the dilatometric and magnetic effects are volume properties, and cannot be conveniently utilized to explore variations in retained austenite from point to point in a suecimen. Some microstructures lend themselves to accurate evaluation of retained austenite contents by metallo-graphic methods,3 but such instances are not too common, and at best, the procedure is tedious. The characteristic X-ray diffraction lines of austenite permit positive identification of its presence; and in the present paper the intensities of these lines, or the corresponding densities* on photographic film, are employed to measure the amount of the retained austenite. In order to cancel the effect of uncontrollable variations in the X - ray exposure and film - development conditions, a standard reference line is Superimposed on each film by maintaining a foil of aluminum in the path of the incident X-ray beam during the entire exposure. With different samples of steel and the same foil, the ratio of the density of some one austenite line to the density of

Jan 1, 1943

By George A. Roberts, Robert F. Mehl

When austenite first forms from aggregates of cementite and ferrite, it is not homogeneous.' This inhomogeneity, consisting of both undissolved carbide and carbon concentration gradients, has a profound effect upon the characteristics of the austenite-pearlite reaction. The rate of formation of pearlite is increased by the presence of inhomogeneities2,3 and because of this (when the inhomogeneity is sufficiently pronounced) the hardenability of the steel is reduced.4,5 In addition to this kinetic effect, inhomogeneities will cause the direct formation of spheroidite from austenite at temperatures just below the Ael.6-11 The present report attempts quantitatively to evaluate the effect of inhomogeneity upon the rate of the austenite-pearlite reaction, presents a method of determining the austenitizing cycle necessary to obtain "practical homogenization," and considers the effect of inhomogeneity on hardenability. Effect of Inhomogeneities on Reactions Near Ae1 Few quantitative data are available representing the effect of undissolved carbide or of carbon concentration gradients upon the rate of the austenite-pearlite reaction. In procuring such data it is necessary to separate the effect of undissolved carbide from that of grain size, and this is not an easy task. The solution of carbide is always accompanied by an increase in the grain size, both effects contributing to the retardation of the reaction rate. It has been possible to make this separation and to obtain data that are independent of the grain size by the use of steels that differ only with respect to their tendency toward grain-coarsening. Steels C and D in Table I are such steels. These steels have been studied previously ;l,2,12 they differ only with respect to deoxidation practice; C was deoxidized with silicon while D, teemed from a portion of the same heat, was deoxidized with additional auantities of aluminum in the mold. C therefore is what is normally termed a "coarse-grained steel" and D is a, "fine-grained steel." These steels can be compared at the same grain size by heating c to a low temperature and heating D to an elevated temperature, producing a structure of austenite plus residual carbide in the former while forming homogeneous austenite in the latter; in such

Jan 1, 1943

By Carl A Zapffe, George A. Moore

In a previous publication from this laboratory1 the conclusion was drawn that the embrittling effect of occluded hydrogen on iron and steel must result from the precipitation of the gas within small openings through the crystal structure. A review of the literature then published indicated the real existence of crystal substructures of the mosaic type and gave presumptive evidence that the "rift openings" associated with hydrogen occlusion are the same as the L'disjunctions" of the mosaic theory. Inasmuch as no clear conception of the actual nature of the crystal fragments and openings inherent in these theories has been established, it was considered desirable to obtain direct micrographic evidence of their true configuration. In this direction the standard methods of metallographic attack have proved useless. Recourse was had, therefore, to the direct examination of the surfaces upon which failure by cleavage had resulted from the presence of hydrogen. It was expected that such a surface, unaffected by chemical attack and undis-torted by polishing, would reveal the fine openings penetrated by the gas and thus indicate the real nature of the substructures. Materials Studied In commercial steels, defects caused by hydrogen take various external forms according to such factors as structure and purity, and to the extent to which hydrogen either occludes or produces chemical changes. The present work relates only to the well-known transcrystalline hydrogen embrittlement—-"pickling brittleness "—in which elemental hydrogen occluded within the grain develops cleavage characteristics. For example, hydrogen embrittlement resulting from hydrogen reaction products is excluded from detailed study. It was also desirable, as far as possible, to have the study uncomplicated by other elements which themselves cause brittleness. Three types of iron were used. These included ingot iron, decarburized free-machining steel, and a specially prepared, aluminum-killed, electrolytic iron. This choice allowed a comprehensive selection of unavoidable minor impurities and inclusions, which could subsequently be shown to have no specific effects on the structures within the grains. Carbon, and any other impurities free to react with hydrogen, were first removed by a preliminary annealing treatment in hydrogen. Subsequent treatment in vacuo extracted much of this hydrogen and its gaseous reaction products. The resultant ductile iron was then embrittled either by a second anneal in hydrogen, followed by rapid cooling, or by cathodic electrolysis. Group 1: Prepared from Armco Ingot Iron As shown in Table 1, the principal impurity in this iron was oxygen, which survived the purifying treatment to a large extent. Two series were prepared. The first (Nos. SIO-24) was purified only one day in hydrogen and dehydrogenized one day in nitrogen. Containing both hydrogen and its

Jan 1, 1943

By Eric R. Jette, Earl S. Greiner

The constitution of the iron-rich iron-nickel-silicon alloys at 600ºC. has been determined by X-ray crystal structure methods. If the results are represented isothermally by a conventional equilateral triangular plot, the single-phase solid solution (body-centered cubic structure) of the iron-rich alloys at 600°C. is bounded by four two-phase areas. Three three-phase areas separate the respective two-phase areas. Three of the phases in the polyphase areas have the same structures as the corresponding phases in the binary iron-nickel or iron-silicon alloys. There are two phases, however, that have not been reported previously and whose structures have been tentatively identified in the present investigation. The solid solubility limit of the iron-rich solid solution at 600°C. was determined by comparing the lattice constants of this body-centered cubic phase in the single and polyphase areas. It has been shown that the lattice constants of ternary alloys may be represented by surfaces whose distances above the ternary plot (representing atomic composition) are proportional to the lattice constants of the respective alloys. Then the intersections of the surface for the single-phase alloys with those for the corresponding phase in the polyphase alloys, at the respective temperatures, are the solid solubility limits of the single-phase alloys at these temperatures. The data of the present investigation were evaluated by this method, although by a system of orthographic projection it was possible to represent the three-dimensional model in two-dimensional diagrams. This investigation has shown that the solid solubility of nickel in the body-centered cubic phase of the iron-rich iron-nickel-silicon alloys at 600°C. changes from 3.7 weight per cent (Owen and Sulley) in the binary iron-nickel alloys to 5.0 per cent in the ternary alloys containing 4.8 per cent silicon. Further additions of silicon increase the solid solubility limit of nickel to 16.0 per cent at 12.6 per cent silicon. From this point the limit decreases to 9.5 per cent nickel at 15.5 per cent silicon and is, of course, zero at 15.3 per cent silicon in the binary iron-silicon alloys.

Jan 1, 1943

By Walter Crafts, C. M. Offenhauer

In a previous study of the carbide phase of chromium steels, it was shown that chromium carbide (Cr7C3) is 2 more stable carbide than cementite (Fe3C) at tempering temperatures above about 500°C. in quenched steels containing from about I to 7.5 per cent chromium. Below this temperature the carbide phase was found to consist of cementite in all of the steels examined. In the present investigation it was found that a similar change in the carbide phase takes place in molybdenum and in chromium-molybdenum steels at approximately the same tempering temperatures. The approximate ranges of occurrence of the alloy carbides Cr7C3, Cr4C, and Mo2C have been determined in steels containing up to 5 per cent chromium, I per cent molybdenum and 0.60 per cent carbon. The experimental steels were made in small high-frequency furnaces from Armco ingot iron, low-carbon ferrochromium and calcium molybdate and were poured into 2-in. ingots, which were forged to I-in. round bars. Sections of the bars were heated to the carbide-solution temperature and cooled rapidly to induce austenite transformation at relatively low temperatures. The steels were subject to some segregation, therefore comparatively high temperatures were necessary to dissolve all of the carbides. The microstructure of the quenched specimens was examined to verify the belief that the carbides had been dissolved, and also to determine the nature of the nonmartensitic structure of incompletely hardened steels. Ferrite and pseudo-martensite were present in some of the specimens, but all specimens were quenched at such a rate that pearlite was not formed, in order to ensure the formation of the carbides at temperatures lower than about 500°C. It had been found previously1 that in chromium steels the chromium carbide could be produced by transformation of austenite in the pearlite range above 500°C., and, in order to start the tempering treatment with the carbide present as cementite, the absence of pearlite was considered to be essential. The specimens were tempered for 64 hr. to permit the formation of the stable carbide. After tempering, the specimens were submitted to microscopic and X-ray examination. The carbides were separated either electrolytically in a 5 per cent HCl electrolyte or with an acidified copper-potassium chloride solution. The crystal structure of the carbide phases was determined by the Debye-Scherrer type of X-ray patterns. The X-ray patterns obtained from specimens near the boundaries between two types of carbides were generally weak and diffuse as compared with the other patterns. For this reason, and because the more soluble components may be masked, the boundaries between the phases could not be determined precisely. However, the X-ray analysis of electrolytically separated carbides has given consistently reproducible results

Jan 1, 1943