Search Documents

By Julian Szekely

Through the simultaneous statement of Maxwell's equations. the turbulent Navier-Stokes equations and the differential thermal energy balance equation, a mathematical representation has been developed for heat transfer and fluid flow in electric arc furnaces. In broad outline the theoretical predictions appear to be consistent with experimental measurements. A discussion is presented of the general arc furnace control and optimization problem, together with the potential for the incorporation of the mathematical model given in the paper, into future control and optimization schemes. ELECTRIC ARC FURNACE STEELMAKING accounts for about 24% of the total steel produced in the U.S., consuming about 550kw-hr/ton of steel which amounts to about 1.4x1010 kw-hr annually. The percentage of steel produced by the electric arc furnace route has doubled since 1960 and is expected to increase further in the foreseeable future, particularly with the more extensive use of direct reduction technology. Table I shows, in terms of percentages, the principal items that contribute to the cost of running an electric arc furnace. It is seen that in addition to the fixed costs (labor, depreciation, taxes etc.) electrode costs and power consumption are the dominant inputs, and it follows that process changes that would ultimately reduce any of these cost components would have a significant impact on the efficiency of the operation. There are three main areas where progress in arc furnace design and operation must be directed in the next decade (Committee on Technology, 1981). 1. Increased furnace productivity 2. Reduction in processing cost 3. Improvement in the thermal efficiency of the furnace An estimate of the current operating efficiency of the electric arc furnaces may be obtained by comparing the theoretical minimum energy consumption per ton of steel with those achieved in operational practice. Theoretically 330-360 kw-hrs of electricity are required to melt one ton of cold scrap. For superheating, 8-12 kWh per 100°F are required. The actual energy requirements are given in Table II. It is seen that furnaces currently operate at 70-80% efficiency, with the major heat loss from the furnace resulting from the waste gases (Committee on Technology, 1981). While it is apparent that operational experience has produced a technology that is reasonably energy efficient this has been achieved in the almost complete absence of any detailed knowledge concerning the fundamental heat transfer mechanisms occuring in the furnace. The purpose of the present paper is to present a case for the usefulness of mathematical models in increasing the thermal efficiency of furnaces by investigating these heat transfer mechanisms and then combining this knowledge with existing electrode regulator mechanisms to produce truly efficient power cycles for melting and refining. In particular, existing mathematical models describing convection, radiation, and electro-magnetism in the furnace should be employed, and the development of optimization models which balance the cost of electrodes and energy against the thermal efficience of the furnace should be undertaken. Recent work in this laboratory (Szekely, McKelliget, 1981, 1982, 1983; Szekely, McKelliget, Choudhary, 1983) has made a useful start towards solving the problem of heat transfer within the electric arc furnace but the work still needs to be directly related to existing furnace power cycles and tested against plant scale measurements.

Jan 1, 1984

By R. F. Domagala, M. Hansen, D. J. McPherson

On the basis of metallographic analysis, incipient melting data, thermal analysis work, and X-ray diffraction, phase relationships in the 0 to 50 atomic pct alloy content were carefully resolved. Phase relationships in the higher alloy content regions were outlined. A single intermediate phase, peritectically formed, of the form ZrX2 was established in each system. A eutectic and a eutectoid were found in the zirconium-rich region of both systems. Appreciable solubility of molybdenum and wolfram was found in ß zirconium, but only a slight solubility in a zirconium. PREVIOUS work on these systems was limited to X-ray investigations of the intermediate phases1, 2 and preliminary surveys of the zirconium-rich alloys." The diagrams presented were determined principally by metallographic, thermal analysis, and incipient melting techniques. In accordance with the Atomic Energy Commission contract requirements, phase relationships up to 50 atomic pct molybdenum or wolfram were carefully resolved, while only a limited amount of work was done to outline the 50 to 100 atomic pct alloy region. A description and analysis of the metals used in the preparation of alloys for this study are included in Table I. A "low-hafnium" zirconium crystal bar, produced by the decomposition of a volatile iodide onto a hot filament, was employed for these studies. Westinghouse "Grade 3" crystal bar was used for all alloys. The zirconium, as received, was coated with corrosion product from a standard autoclave test by which its grade designation is determined. This was removed by a sand-blasting and HF-HNO, pickling technique developed by the Atomic Energy Commission. The bars were cold rolled to approximately 1/32 in. sheet, pickled for iron removal, sheared to 1/4 in. squares, washed with acetone, and stored for furnace charge. Molybdenum sheet. 0.003 in. thick, 0.001 to 0.005 in. wolfram wire, and M in. wolfram rod crushed to granular form, from Fansteel Metallurgical Corp., were used as alloying stock. Melting Procedure A nonconsumable electrode arc furnace identical to that described in detail by Hansen et al.4 was used. The source of power for melting Zr-Mo alloys was a 400 amp dc welding generator, while a 900 amp dc generator was used for preparing Zr-W alloys. Electrode tips of the same material as the alloy addition were employed in both systems, i.e., molybdenum tips were used for Zr-Mo alloys, and wolfram tips for Zr-W alloys. This precluded any possible electrode contamination during arc melting. Ingots of 20 to 30 g of Zr-Mo alloys were melted under high purity helium in the cavity of a copper block insert in a spun copper crucible. Homogeneity was insured by remelting all ingots four to six times without opening the furnace. Due to the large difference between the melting points of wolfram and zirconium, direct melting of the charged metals as above generally yielded ingots containing unmelted wolfram particles. This was overcome, after many variations in technique, by melting a pure wolfram charge at high power levels (700 to 900 amp) and slowly adding zirconium to form a master alloy. Such master alloys were sectioned and remelted several times, then crushed to serve as melting stock for alloys of lower wolfram content. Master alloys containing 26, 50, 52, 70, and

Jan 1, 1954

By W. David Leyshon

This paper compares the general characteristics of Florida phosphate rock for the manufacture of fertilizers and phosphoric acid. The subjects covered are relative reactivity for superphosphate manufacture and various properties of the rock for use in phosphoric acid production, including gypsum filterability, sulfuric acid consumption, foaming, relative cor-rosiveness, chemical composition, and others. Comprison will include rocks of North Africa, the western U.S., and other locations. This paper reviews some of the important properties of phosphate rock, giving particular attention to a comparison of Florida phosphate with some of the major phosphates produced in the world. It also discusses some of the techniques available for evaluatine Phosphate. rock for phosphoric. acid and fertilizer use. The use of Florida rock outside the U.S. has been increasing steadily, evidence in itself of certain desirable properties of this rock. Table I is a list of the major producers and exporters of phosphate rock as of 1965. Florida, meaning in this case Central Florida, was the largest producer of phosphate rock and the largest exporter of rock. Of a total world production estimated at 60 million tons, about 50 million tons were produced during 1965 by the areas listed in Table I. One major producing area omitted from Table I and from subsequent evaluations in this paper comprises Nauru, Ocean, and Christmas Islands, whose output totaled about three million tons in 1965. This phosphate is marketed entirely by the British Phosphate Commissioners, primarily in Australia and New Zealand, and does not get into the world market elsewhere to a significant degree at the present time. South Africa, Egypt, Israel, and North Vietnam also produce lesser quantities of rock for the international market, and these materials are also excluded from this review. The phosphate industry is now seeing the introduction of several new areas that will be significant factors in the rock market in the future. The North Carolina and North Florida areas each have capacity at present or in the near future for producing about three million tons per year. The Djebel Onk mine in Algeria has a potential of about 800,000 tons per year. In addition, announcements have been made of significant development of deposits in Peru and in the Spanish Sahara, and sizable expansions are planned in several of the existing areas that are now of minor importance. Competition for the rock market appears to be increasing so that more attention will probably be paid to evaluating phosphates available from different areas. It is not possible here to go into all the properties of even a few of the materials listed in Table I. However, it is practical to discuss some comparative properties important for making phosphoric acid and fertilizers, and to touch on the techniques used for the evaluation of various phosphate rocks. Table II gives typical analyses of the major phosphate rocks available in the world as of 1965. There is, of course, an inherent danger in labeling any analysis as typical, but it appears that variations between two samples of rock within one given area are generally less than between two samples from different areas. The analyses listed were compiled from several years of data so they may not be representative of the present output of any particular area. The range of grades represented by these analyses goes from 66 BPL for Gafsa rock to 85 BPL for Kola

Jan 1, 1968

By B. C. Craft, Murray F. Hawkins, John R. Rayne

A reservoir mechanism of sulfur recovery by the Frasch process is presented. Improving the economic, of recovery appears to be largely a well, rather than a reservoir problem. A most important factor is the limitation of the lateral extraction of .sulfur about the wells due to the almost vertical .flow of the injected hot water. Model studies are described which confirin the mechanism of sulfur recovery by presently used methods. Studies on a five-spot system indicate improved econo171ics regarding recovery, recovery rate, therinal efficiency, and well spacing. INTRODUCTION Sulfur mining by the Frasch process is one of the most important methods of sulfur production. Developed by Herman Frasch in 1894, it enjoyed a rapid growth and presently supplies 40 per cent of the world's sulfur and about 80 per cent of the U. S. sulfur." The Frasch process of mining the underground native sulfur deposits involves melting the sulfur in place by introducing super-heated water into the sulfur bearing formations, and producing the sulfur as a liquid. Many problems are encountered in mining these native sulfur deposits, and sulfur production depends chiefly upon the nature of the sulfur deposit and the underground thermal efficiency of the heat transfer system. THE RECOVERY MECHANISM Sulfur is typically found in salt dome cap rock as crystal aggregates occurring within the cavities, seams, and pores of limestone and gypsum formations.' Sulfur-bearing limestone is typically cavernous, vugular, and fractured. Porosity averages about 20 per cent and sulfur saturation averages approximately 25 per cent by weight.1,15 Vhe sulfur bearing formations are very heterogeneous in nature and physical characteristics vary widely within small distances. Total cap rock thickness may vary from 50 to more than 1,000 ft, and net productive thicknesses from a few feet to a few hundred feet. In most cases the upper portion of the limestone formation is barren of sulfur, varying from 5 to 200 ft in thickness.1,15,33 The immediately overlying sediments and the underlying dense anhydrite formation provide permeability barriers to the flow of water from the cap rock section.'" Sulfur wells are drilled with conventional oilfield drilling rigs. Fig. 1 illustrates the casing program of a completed sulfur well. The cap rock sections are drilled with core bits and continuous core logs are recorded. After coring, the cap rock section is underreamed to a diameter of 9 7/8 in. The bottom of the 8-in. casing is welded closed and rests on the bottom of the drill hole, which extends through the cap rock to a few feet above the bottom of the sulfur-bearing formation. The 8-in. pipe contains 5 ft of 3/4-in. perforations on 6-in. centers and 15 ft of 1-in. perforations on 1-ft centers. The 20-ft perforated section is set opposite the lower 20 ft of the sulfur-bearing formation. The 4-in. sulfur pipe is set inside the 8-in. water line on a seat ring at the top of the 5-ft section of 3/4-in. perforations. The mining water, heated to about 330°F under a pressure of 250 psi, is forced down the annulus between the 8-in. and 441-1. pipes and through the 1-in. perforations to the sulfur-bearing formation. Compressed air, at 400 psi pressure, is supplied through a 1-in. air line near the bottom of the well to provide air-lift of the liquid

Jan 1, 1957

By J. B. Clark, A. S. Yue

PfANN1 has shown that when a cylinder of molten binary alloy freezes directionally, the solute distribution can be described analytically by cg=k(l-g)k-1 [1] in which Cg is the solute concentration in the solid at the point where a fraction g of the original liquid has frozen and Co is the initial solute concentration. The term k is the effective distribution coefficient of the solute and is defined as the ratio Cx/CL where Cl is the solute concentration in the main body of the liquid. Both the slope and the intercept at -log (1 - g) = 0 of a logarithmic plot of Eq. [I.] provide a measure of k. Under conditions of freezing where diffusion in the solid is negligible and the solute concentration in the liquid is uniform, the effective distribution coefficient k, derived as above, is equal to the equilibrium distribution coefficient taken from the phase diagram. However, under ordinary freezing conditions, even though solid-state diffusion is negligible, the solute rejected from the freezing solid does not have sufficient time to diffuse uniformly throughout the main body of the liquid. A concentration gradient is formed in a boundary layer in front of the advancing solid-liquid interface as pointed out by Rutter and Chalmers.' Thus, under these conditions, the concentration of the freezing liquid in the boundary layer is higher than that predicted from the equilibrium phase diagram and consequently the effective distribution coefficient k, obtained from directional freezing experiments, is larger than the equilibrium distribution coefficient. The symbol k will designate the effective distribution coefficient in the discussion which follows. In this paper, the determination of the effective distribution coefficient is demonstrated for several dilute binary alloys and a polynary magnesium alloy. The effect of freezing rate on the value of the derived k is presented in detail. An example for the calculation of the solidus in a dilute binary solution using the experimentally determined effective distribution coefficient is given. MATERIALS STUDIED Binary magnesium alloys of copper, nickel, and zinc containing approximately 0.10 wt pct solute and a polynary alloy containing approximately 0.10 wt pct each of several elements were selected for study. The spectrographic analyses of these alloys are given in Table I. The analyses are believed accurate to within +5 pct of the stated value. EXPERIMENTAL PROCEDURE Extruded rods of the alloys, 3/8 in. in diam and 11 in. long were used for directional freezing studies. The alloy rods were melted horizontally in a high-purity graphite boat by a Nichrome-wound tube furnace which moved parallel to the length of the rod. The boat was supported inside a Vycor tube suitably connected to provide a protective atmosphere of dry SO2. For an experimental run, the following procedure was used. The alloy rod was cleaned in dilute HNO, and dried. The graphite boat containing the alloy rod was placed within the Vycor tube and the tube connected to the gas train' The 'ystem was purged for 20 min with dry argon and then a flow of dried SO, was maintained for the duration of the run. The movable furnace was positioned and either all or a desired portion of the alloy rod was melted. After attainment of steady-state temperature gradients the molten bar was solidified from one end to the other at a controlled rate by pulling the furnace along the length of the bar at the desired speed. At the conclusion of the run, the directionally frozen alloy rod was analyzed spectrographically at 1 1/4-in. intervals.

Jan 1, 1961

By W. I. Duvall

A brief review of the resonant frequency and ultrasonic pulse methods for obtaining elastic constants of rock samples shows that the equations for an iso-tropic elastic solid commonly used to calculate Young's modulus, modulus of rigidity and Poisson's ratio can result in appreciable error if the rock is even slightly anisotropic. Consideration of the equations that relate bar velocities and free medium velocities to elastic constants for orthotropic, transversely isotropic and isotropic solids shows that a combination of the two methods can be used to determine average elastic constants for a single sample of rock. The determination of elastic constants in different directions requires a large number of tests on oriented samples. The use of both the resonant frequency and ultrasonic pulse methods is recommended so that the determination of both Young's modulus and modulus of rigidity is independent of the determination of Poisson's ratio. The dynamic elastic constants of rock can be determined in the laboratory from measurements on small samples which may be in the form of right circular cylinders (diamond drill core). Either the resonant frequency6 or ultrasonic pulse method1 can be used to determine the longitudinal and shear velocities in the same specimen. The resonant frequency method provides a measure of the longitudinal bar velocity, V,, and the torsional bar velocity Vt; whereas the ultrasonic pulse method provides a measure of the free medium longitudinal velocity, Vp, and the free medium shear velocity, V, . The elastic constants (Young's modulus E, modulus of rigidity G, and Poisson's ratio v) are calculated from the density of the specimen, p, and one of the pairs of velocities by means of equations from isotropic elastic theory. For the resonant frequency method the equations are: It should be noted that Eq. 5 is derived from Eq. 4 and 6 and the relation between the elastic constants for an isotropic solid as given by the first part of Eq. 3. The in situ determination of dynamic elastic constants is accomplished by generating a complex seismic pulse in the rock and measuring the arrival times for the longitudinal and shear phases of the pulse at two distant points.5 From these travel times and the distance of the travel path both Vp and Vs are determined. The density of the rock is determined from samples in the laboratory, and Eq. 4 to 6 are used to calculate the elastic constants. These three methods involve the assumption of isotropy and perfect elasticity, and require the use of the relation v = (E/2G) - 1, which holds only if the elastic constants are the same in at least two directions. The calculation of v by means of Eq. 3 introduces considerable error, as a 1% error in V, and V, results in a 20% error in v. If frequencies for samples with a length-to-diameter ratio greater than four and a length greater than 8 in. are measured carefully, the experimental error can be kept to 1%. However, it is not unreasonable to expect that the variation in Vb and Vt with direction may exceed 1%. Thus, for even a small percentage of anisotropy the use of Eq. 3 can result in considerable error in v. The calculation of v by Eq. 5 also involves some uncertainty as a 1% error in Vs and Vp can result in a 6% error in v. By means of the pulse method both

Jan 1, 1965

By N. S. Bornstein, M. A. DeCrescente

ThE nickel and cobalt base superalloys employed in gas turbine engines under some conditions' oxidize at an accelerated rate when exposed at elevated temperatures to atmospheres containing minor quantities of sea salt. It is generally agreed by investigators in the field that the major steps comprising this phenomena of accelerated oxidation, also called hot corrosion or more commonly, sulfidation, are a) a slag composed primarily of Na2S04 is formed on turbine components, b) the protective oxide scale is rendered ineffectual, c) the slag reacts with. the substrate to form metal sulfides, and d) the substrate is rapidly oxidized. Sodium sulfate, a minor constituent of sea salt, is formed within the gas turbine engine from the reactions between NaC1, the major constituent of sea salt, and the oxides of sulfur.' Depending upon concentration, engine compression ratio, and temperature, the Na2S04 will be present in the gaseous or condensed state. It has been determined that in order for sulfi-dation to occur the Na2S04 must be present in the condensed state.' Although a number of sulfidation mechanisms have been postulated, there is no consistent explanation of oxide scale breakdown and subsequent catastrophic oxidation. Some investigators believe that after oxide scale breakdown, the Na2SO4 reacts with the substrate to form lower-valent sulfur compounds. The accelerated rate of oxidation is due to oxidation of a nickel-sulfide phase." Other investigators believe that sulfur from the fused salt reacts with chromium in the alloy to form chromium-rich sulfides; thereby depleting the alloy of an element which imparts oxidation resistance.3,4 This publication will demonstrate that the accelerated rate of oxidation associated with sul-fidation is due to interactions between Na20, a con-stituent of NazSO.,, ancl the substrate. The aim of this study was to separate and determine the effect of compounds of sulfur and sodium on the oxidation kinetics of various nickel-base superalloys, in order to elucidate the sulfidation mechanism. The work reported herein is sponsored in part by the Naval Ship and Development Center-Investigation of EXPERIMENTAL PROCEDURE The nickel base alloys B-1900, U-700, and Waspaloy were chosen for this study. The nominal compositions are presented in Table I. The specimens, approximately 1 in. by 1 in. by 0.060 in. received in the "as cast" condition were polished with 600 grit emery, washed, and rinsed with acetone immediately prior to oxidation experiments. Sulfida-tion attack was duplicated in the laboratory by apply-ing a saturated aqueous solution of Na2S04 onto the al loy substrate and subsequently exposing the specimens in oxygen at elevated temperature. The method of Na2S04 application is described elsewhere in detail.6 In another set of experiments the affected substrate microstructure associated with sulfidation was reproduced free of the normal oxide scale, by reacting and diffusing elemental sulfur into the alloy. The experimental procedure is also described elsewhere in detail. All oxidation studies were performed using an Ainsworth thermobalance type RV-AU-1, which is readable to 0.01 mg and is reproducible to ±0.03 mg. The specimens, introduced into a heated furnace, were exposed in the temperature range 800" to 1000°C

Jan 1, 1970

By C. Elbaum

Specimens consisting of several crystals, each with a prescribed geometry and a controlled orientation with respect to both the extermally intposed stress and the highboring crystals, werer deformed in tension at room temperature. It is shown that the influence of a grain boundary on the plastic defolrmation of a mullicrystal depends on the extent to which the presence of the grain boundary, contributes to the necessity 01. multiple slip in the component crystals. MANY investigations were carried out in the past on the plastic deformation of aluminum single crystals, bicrystals, and polycrystals. The various studies of bicrystals aimed primarily at elucidating the influence of grain boundaries on the mechanisms of plastic deformation. In the earlier studies the emphasis was placed primarily on finding the "strength" contributed by a grain boundary. In more recent investigations it became apparent that the grain boundary does not contribute any strength per se but rather acts by influencing the mechanism of plastic deformation in the adjoining grains. The concept thus emerged of the plastic properties of a bi-crystal, and by extension of a polycrystal. depending solely on the mechanism of deformation prevailing in each grain. The contribution of the grain boundaries thus becomes limited to the type of restrictions imposed on each grain by its neighbors. The present investigation of the behavior of specimens consisting of more than two crystals was designed for the purpose of studying the influence of the grain boundary in simple aggregates. The findings of this study further support the concept that the restriction imposed on a deforming crystal by its neighbor constitutes the predominant factor in the plastic deformation of polycrystals. The term "multicrystal", in contrast with single and polycrystals, is used to describe specimens consisting of several crystals. Each one of the component crystals has a prescribed geometry and a controlled orientation with respect to both the externally imposed stress and the neighboring crystals. Two groups of multicrystals were prepared. All the component crystals of both groups of multicrystals had the axial orientation (isoaxial multicrystals) shown in Fig. 2, with the primary slip plane and slip direction both at 45 deg to the specimen axis. In the first group (hereafter referred to as group I) the component crystals were symmetrically oriented with respect to the plane of the grain boundary; the primary slip plane and slip direction were respectively at an angle of 60 and 30 deg to this plane. For the second group (group 11) the axial orientation remained the same and the relative positions of the primary slip plane and slip direction are shown in Fig. 3. EXPERIMENTAL METHODS The material used was aluminum 99.993 pct pure, according to the manufacturer. The specimens were produced by growth from the melt in a horizontal graphite container. In order to maintain the prescribed geometry of the individual crystals of a multicrystal, a technique described elsewhere was used.' The specimens were grown under a small, positive pressure of helium at a rate of approximately 1 mm per min. annealed in air for 24 hr at 640°C and cooled over a period of several hours to room temperature. All the multi-crystals were rectangular in cross-section, 6 by 18 mm (see Fig. 1) and from 150 to 170 mm long. In each group bicrystals, tricrystals and quadru-crystals were prepared as shown in Figs. 1 and 4.

Jan 1, 1961

By Donald Ofte, L. J. Wittenberg

The absolute viscosity coefficients and activation energies for viscous flow of bismuth, lead, and zinc are reported. The viscosities were measured in an oscillating cup viscosimeter from nearly 1000°C to within 1 deg of the melting point of zinc; 2 deg of the melting point of lead; and in the supercooled condition of bismuth, 9 deg below the melting point. Two methods of relating the change in the viscosity coeflicient with temperature were compared. From the plot of the logarithm of viscosity (77) as a function of reciprocal absolute temperature (1/T), the equations of the lines which fit the data are: AS a part of the Reactor Fuels and Materials Research program at Mound Laboratory, precise measurements of the physical properties of liquid metals are of interest because the desirability for a low vapor pressure liquid coolant in high-temperature nuclear power plants has focused attention on the properties of the liquid metals.' Also, a clearer understanding of the liquid metal state can be achieved when better measurements are available. For these reasons, the viscosity values of certain liquid metals have been determined. The viscosities of the liquid metals, bismuth, lead, and zinc are measured in an oscillating cup viscosi-meter. In this apparatus the liquid is sealed in a right-circular cylindrical cup which is attached to a torsion fiber so that the system forms a torsion pendulum. The damping effect of the molten metal upon the normal oscillations of the torsion pendulum is a function of the viscosity of the liquid. This method is advantageous because of the small amount of sample required (8 to 10 cc), the ease of remote operation of the apparatus in a vacuum system, and the containment of the liquid metal in a hermetically sealed, nonreactive, tantalum crucible. With the liquid metal sealed in the crucible, loss of liquid by vaporization and chemical reactions with atmospheric gases are prevented. Additionally, an absolute viscosity determination is obtained which avoids the sources of possible error in a relative measurement. The difficulties in a relative method include the repetitious cup calibrations, and the uncertainty inherent in the practice of calibration of the apparatus at room temperature and extrapolation to the high temperatures necessary for the investigation of most liquid metals. The oscillating cup viscosimeter was first described by Meyer in 1891.2 The first complete mathematical treatment offering an absolute measurement of viscosity with this type of apparatus was proposed by Andrade and chiong3 in 1936 for an oscillating sphere filled with a liquid. This solution was modified to include a right circular cylinder by Hopkins and Toye4 in 1950, and R. Roscoe5 in 1958. Roscoe's treatment for the calculation of the viscosity of a liquid at a particular temperature from the logarithmic decrement and period of oscillation of the torsion pendulum was used in this investigation. A detailed description of the method of calculating the viscosity from these measurements has been reported earlier.' This calculation has been programmed for the IBM 704 and IBM 1620 computers for expedition of the otherwise laborious calculations. APPARATUS The cutaway view of the apparatus in Fig. 1 shows the suspension system within the high vacuum enclosure. The low residual gas pressure. maintained between 1 x 10-4 and 5 X 10-6 mm Hg during a determination, avoided both air drag on the pendulum and

Jan 1, 1963

By R. E. Maringer, B. A. Wilcox, A. H. Clauer, J. H. Sovik

Internal friction measurements from 25° to 1100°C have been made on specimens machined from TD nickel bar. The use of miniature torsional pendulum specimens alloived damping measurements to be made on specimens whose axes were parallel to the original bar axis, at 45 deg to the bar axis, and at 90 deg to the bar axis. There was a marked difference in damping with varying specimen orientation, and this was related to the pronounced micro structural anisotropy of this material. Varying the microstructure by cold working plus re crystallization annealing also greatly affected the damping behavior. Recrystallization essentially removed the broad peak found at 400° to 650°C in as-received TD nickel, and in addition greatly lowered the background damping. DlSPERSION hardened metals such as SAP-type alloys (Al + AI2O3) and TD nickel (Ni + 2 pet ThO2) have been developed primarily to improve high temperature strength properties. There have been numerous damping studies on SAP-type alloys,1"9 and one of these6 showed that the damping capacity was greater than that of 99.5 pet pure Al from 25° to 600°C. Several preliminary observations on TD nickel10 and an experimental Ni-ThO2 alloy11 indicated that here, too, the damping capacity was quite high. A high damping capacity in TD nickel, combined with its relatively good high temperature creep strength, could provide an additional benefit for some high temperature structural applications. This work was undertaken to study the internal friction behavior of TD nickel bar over the temperature range 25° to 1100°C. In Ni-ThO2 and Ni-Cr-ThOz alloys, other properties such as steady-state creep rate,12 creep rupture life,12'13 modulus of elasticity,14 and ultimate tensile strength,14 are sensitive to orientation with respect to the primary working direction. Thus, in order to examine anisotropy effects in damping, studies were made on specimens machined with their axes parallel, at 45 deg, and transverse to the original bar axis. MATERIAL DESCRIPTION AND EXPERIMENTAL PROCEDURES The material used in this investigation was from the same £ in. diam TD nickel bar that has been studied previously.13'15 The structure typical of the as-received TD nickel is shown in Fig. 1, and additional details of the structure can be found in Refs. 13 and 15. Elongated grains are parallel to the bar axis and are about 10 to 15 u long and about 1 u in diam. This structure is very stable, since no recrys-tallization or grain growth occurred after annealing for 3 hr at 1300°C.13 The total ThO2 content was 2.3 vol pct and the particles ranged in diameter from 100 to 1000. The average particle diameter was 370Å and the mean planar center-to-center particle spacing was 2340Å. To examine the influence of grain and subgrain structure on damping, a section of TD nickel bar was swaged and drawn at room temperature from 1/2 in. diam bar to 0.040 in. diam wire (99.3 pct re-duction in area). The structure in Fig. 2 resulted after the wire was annealed for 3 hr at 1250°C. The recrystallized grains were much larger than the grains in as-received bar, but still showed considerable elongation in the direction of working. A high

Jan 1, 1970

By C. F. Floe, M. Cohen, M. B. Bever, V. G. Paranjpe

NITROGEN is becoming recognized as one of the important elements in ferrous physical metallurgy. Several investigations indicate that nitrogen plays a significant part in such phenomena as strain aging and temper brittleness. Nitrogen is known to be a powerful austenite stabilizer and may be a useful alloying element in stainless steels. The hard-enability of steels has been found to be increased by nitrogen. The nitriding and carbonitriding pro- cesses are examples of successful industrial application of nitrogen in ferrous materials. For a complete understanding of the various roles played by nitrogen in ferrous metallurgy a knowledge of the phase relations in the binary iron-nitrogen system is essential. The literature contains a variety of iron-nitrogen phase diagrams. Those suggested by Fry, Sawyer,² Epstein and his coworkers,³,4 Lehrer,5 Eisenhut and Kaupp, and Hägg7,8 differ significantly from one another. Moreover, the most recent of these investigations dates back about twenty years. A re-determination of the entire diagram, therefore, appeared desirable. In addition to the published phase diagrams, investigations have been carried out on portions of the iron-nitrogen system. Bramley and Haywood determined the composition and temperature of the iron-iron nitride eutectoid. The maximum solubility of nitrogen in a iron was investigated by Portevin and Seferian10 and by Diikstra. 11 Several investigators have determined the crystal structure of the various iron-nitrogen phases. In fact, Eisenhut and Kaupp6 and Hägg7 based their entire phase diagrams on the results of such studies. Epstein' and Emmett12 reported lattice-parameter measurements on the E (hexagonal close-packed) phase. Osawa and Iwaisumi13 also studied the change in lattice parameters of the a, ? and E phases with nitrogen content. Brill" and Hägg7 showed that the y' phase has an ordered structure, and Hendricks and Kostingsl5 suggested a structural relation between the E and phases. In a recent investigation Jackl6 arefully studied the super-lattice reflections from the y' and phases. Experimental Procedure A large variety of experimental methods has been used by previous investigators in studying the iron-nitrogen system. These methods include thermal analysis, microscopic examination, X ray diffraction, magnetic measurements, dilatometric analysis and anelastic experiments. Thermal analysis is of low accuracy in determining changes in the solid state. This accuracy is further decreased for iron-nitrogen alloys since composition changes will occur during heating or cooling unless extreme care is taken to use a proper atmosphere. Metallographic methods are of small value . because homogeneous iron-nitrogen alloys of sufficient size can be produced only with great difficulty. The dilatometric method is open to the same objections as thermal analysis in that changes in composition may occur during heating and cooling. Lattice parameter measurements by X ray diffraction can be used satisfactorily for determining phase boundaries if cooling is fast enough to prevent structural changes. In view of the limitations of the various other techniques two methods only were adopted in this investigation, (1) X ray diffraction measurements, and (2) a controlled nitrogenizing method. The second method is an innovation for investigating the iron-nitrogen system, and will be described at length. Materials: All iron-nitrogen alloys were prepared from carbonyl iron powder of the analysis shown in table I. The average particle size was about 20

Jan 1, 1951

By M. C. Fuerstenau, A. M. Gaudin

A homogeneous suspension is viewed by X-rays. The radiation density seen, affords a measure of the extra absorption due to the solids contained. This radiation density, at a predetermined depth, varies with time. The chart of this relationship gives the desired determination. s it is frequently necessary to know the distribu-rt tion of particle sizes in various products used in industrial applications, a number of means have been devised for these determinations. Some of them include direct microscopic measurement, screening of product, extraction of samples from a settling slurry, air and liquid elutriation, adsorption of gases and liquids, and light absorption. The absorption of light as a means for particle size analysis has been known and employed for some time. X-ray absorption has not received the attention that light has enjoyed primarily because of the complex and costly instrumentation involved with X-rays. In 1954, Brown and Skrebowskil suggested the use of X-rays for this type of determination but did not supplement their suggestion with experiment. In 1958, the authors conducted size analysis experiments on suspensions of quartz with the Trans-viewer: an X-ray unit specifically adapted for sedimentation research, and Rossa conducted experiments in a similar manner in 1959 with a stable gamma emitter, AmZ4l, on suspensions of uranium oxide. To illustrate the applicability of X-ray absorption for analyzing particle size distributions, an essentially mono-energetic and well collimated X-ray beam was used as a source. The beam was made mono-energetic by placing a zirconium filter before the counting window together with feeding the pulses from the proportional counter into a reverter, a device capable of discriminating against pulses of various heights. The sketch in Fig. 1 shows the collimation arrangement in more detail. As shown, the X-ray beam was collimated by inserting a 3/8-in. ID pipe into one of the ports of the X-ray tube and further by two small steel plates with slits 3/32-in. wide and 1/2-in. long placed in front of the window of the Geiger-Mueller (G-M) tube. One of the plates was located 1/2 in. and the other 2 1/2 in. from the window of the counter; both plates were contained in an aluminum holder mounted on the box containing the G-M tube. This type and degree of collimation should eliminate any intensity due to scattered radiation; the transmitted intensity would then be solely a functior. of the absorption by the suspension at that level. Experimentally, homogeneous suspensions of finely ground quartz were prepared, Daxad (a sulfo-nated compound made from wood) was added as dis-persant, and the transmitted X-ray intensity measured as a function of time at a fixed depth in the suspension. In more detail, the experimental procedure involved the following: 1) measuring the transmitted X-ray intensity through the Daxad solution at a predetermined position or settling depth in the absence of solids, 2) placing a known amount of dry solids into the vessel and agitating thoroughly, 3) measuring the transmitted X-ray intensity at time zero (i.e., immediately upon cessation of stirring), and 4) measuring the transmitted X-ray intensity as a function of time (a 10-sec counting interval being chosen and counting being started 5 sec before the desired time). Since a linear relationship exists between percentage solid in a slurry and the logarithm of transmitted intensity for a homogeneous material, the cumu-

Jan 1, 1961

By H. C. Rogers, B. W. Roberts

Vanadium foils and wires, either cold-worked or recrystallized, show a ductile-brittle-ductile fracture sequence with temperature. At about 150°C the hydrogenated vanadium wires are found to be ductile. At room temperature the wires show brittle fracture, and at liquid nitrogen temperature the wires are again ductile, although to a lesser extent than found at 150°C. Hydrogenated cold-rolled vanadium foils have been found to fracture in unusually straight lines along the [100] texture directions. The hydrogen content required to show this [loo] texture fracture is limited to a few atomic percent in range. Hydrogen content above this range causes fracture to occur in random directions. PRELIMINARY experiments with hydrogenated vanadium foils indicated that the embrittling effect of the hydrogen on the vanadium, which was severe at room temperature, was reduced at both lower and higher temperatures. In order to investigate more quantitatively the ductility changes with temperature, vanadium wires were tested in tension at three temperatures. The reduction of area was measured and the fractures examined. Specimen Preparation—The material used was kindly supplied by A. U. Seyboltl in the form of M-in. rod. The nominal analysis was: 99.2 pct V assay, 0.2 pct C, 0.01 pct Si, 0.01 pct Fe, 0.015 pct Ca, 0.02 pct O,, 0.01 pct N2, and 0.003 pct H,. The specimens were 0.035-in. diam wires. These were prepared by cold-swaging the supplied rod to the final diameter without intermediate anneal. They were finished with long-bearing dies. The surface was then given a light polishing with 0 emery paper and degreased in trichlorethylene. An X-ray check indicated that the cold-swaged wire had a fairly strong [110] fiber texture. The wires were then recrystallized in vacuum in continuously pumped quartz tubes at two temperatures to give two different grain sizes. One group of wires was given 45 min at 800°C; another, 60 min at 1000°C. The microstructures in Fig. 1 show the wires after they have been electropolished and elec-troetched in 10 pct HCI. After the 800°C anneal, the vanadium grains had a diameter of 1 to 2p and there appeared to be a tendency to elongation in the longitudinal direction. However, the latter may be an etching effect. After the 1000°C anneal, the grains appeared to be equi-axed, but there was a very large spread in grain size, probably due to some impurity segregation in the wire, Fig. 1, right. The grain diameters varied between the limits of 1 to 150p. The small particles shown particularly clearly in Fig. 1, left, are also visible on the polished surface of all samples before etching and are undoubtedly carbides and nitrides of vanadium. An X-ray check indicated these annealed wires had a strong [110} recrystallization fiber texture. The hydrogenation was carried out in an iron retort with dry hydrogen flowing through it, the retort and contents being heated in a tube furnace with a hydrogen atmosphere. The hydrogenation temperature was approximately 500°C. The hydro-genation times chosen were approximately those which would be just enough to cause the vanadium to be brittle at room temperature and thus allow the greatest chance for ductility to occur at the other two temperatures. Part of the fine grained specimens were hydrogenated for 21/2 min at 490°C. Part of the coarse grained specimens were hydro-genated for 11/2 min at 515°C.

Jan 1, 1957

By J. G. Richardson, F. M. Perkins

Results and procedures are pre-cented covering a laboratory investigation of the effect of rate of water advance on the displacement of oil from clean water-wet sands. The experiments included water floods at atmospheric pressure with and without gas present over a range of flow rates from reservoir rates to rates ten thousand-fold higher. Water floods were also conducted at various rates at elevated pressure with no gas present. Other tests were made at elevated pressures by forming a free gas saturation by evolution of gas from solution in the oil before water flooding. Recovery of oil was found to be independent of the rate of water advance with or without gas present and independent of the pressure level of the water flood. A dimensionally-scaled model of a reservoir composed of a thick uniform sand was used to study the effect of rate on the tendency of water to underrun the oil because of the greater density of the water. Another model was used to study the effect of rate on the tendency of water to channel down the more permeable sand in a stratified reservoir composed of sands of diffierent per-meability in intimate contact. The effect of the presence of gas on the flooding conformance was studied with both models. The results obtained using these simple models provide a basis for predicting behavior in more complicated systems. INTRODUCTION In spite of the great amount of research on the flow of fluids in porous rocks, there still exists some disagreement as to the influence of the rate of water advance on the displacement of oil by water flooding. Various investigators have found that increasing the rate of water injection would decrease, increase, or have no effect on recovery of oil. One reason for this disagreement has been a lack of understanding of the effects of the capillary forces which influence the results obtained in the laboratory at low rates with short columns. Other causes of the disagreement have undoubtedly been inconstancy and uncertainty of the wetting properties of the porous media studied, inaccurate saturation determinations, and in some instances conclusions based on insufficient data. Capillary forces are involved in the displacement of oil by water in two ways. On the microscopic scale, the capillary forces control the dis- tribution of oil and water in the pore spaces of the sand. The magnitude of these forces is governed by the preference of the sand to be wet by water, the pore sizes and geometrical shapes, and the interfacial tension between the water and oil. On the macroscopic scale, differences in saturation at the flooding "front" when water displaces oil result in capillary pressure gradients. These gradients cause water to flow preferentially ahead of the front. In short laboratory columns at low rates, the front can span the entire column length. Also, at the outflow boundary the preference of sand to be wetted by water results in retention of water and a high water saturation in the region of the outflow face. This effect is commonly called the boundary effect.' The capillary pressure gradients which complicate the interpretation of flooding performance of short laboratory columns at low rates of water advance become negligibly small under certain conditions. The conditions where capillary pressure gradients are negligible are in long, homogeneous reservoir sands,' at high rates of water advance in short laboratory columns,3 and at any rate in laboratory columns as the residual oil saturation is approached. Thus, laboratory water floods, conducted

Jan 1, 1958

By T. Gleason

Problems involved in applying wet scrubbers to gas cleaning coal dust from thermal dryers are reviewed. Careful consideration of all the elements going into a modem coal preparation plant is required in view of the increasing difficulty of applying suitable gas cleaning equipment. This problem arises not only from the presence of finer coals to be handled, but also from the pressures applied by both the general population and the plant operators to improve the standard of gas cleaning performance. The Peabody scrubber is presented by the author as a machine designed to help improve this cleaning operation. Wet scrubbing of fine coal from thermal dryers poses a number of problems in adapting the gas cleaning equipment to the fine particulate matter. Among them are: 1) Dust in gases leaving the thermal dryer cyclones is often extremely fine with high percentages of total dust loading less than 1p in size. 2) Flue gas entering the scrubber is corrosive since the coal fired in the thermal dryer furnace is in the range of 0.7% by weight sulfur or higher. Therefore, comparatively high cost stainless steel alloys or corrosion resistant linings are required for a satisfactory service life. 3) Requirements for residual coal loadings leaving the gas scrubber are continually being lowered, while modern coal preparation operation is imposing heavier burdens on the collection system through introduction into the thermal dryer system of very fine coal previously not recovered. The advent of flotation systems in increasing numbers is typical of this trend. 4) Installation of a scrubbing system on thermal dryers is made primarily to reduce air pollution and product recovery value is of little or no value. Therefore, the operating costs of gas pressure drop and water requirements must be held to a minimum. We have, then, the basic situation which calls for comparatively high cost stainless steel alloy wet scrubbers, which must be highly effective in removing coal fines particularly in the 0 to 5-p size range. Further, the standards of scrubber performance are continually being raised towards the level where gas discharged from the scrubbers must be essentially free of any coloration due to residual dust. In view of the highly refractive nature of fine coal dust, residual loadings of 0.01 grains per cu ft or less are required to provide a color free discharge gas. In the present specifications under consideration a residual coal loading of 0.05 grains per cu ft would be considered color free, since the discharge gases are saturated in the range of 150°F. The steam plume acts as a masking agent at the point of discharge, and upon dissipation of the 150° saturated vapor stream, the effluent gas stream could be essentially color free. The specification under consideration calls for residual solids loading of 0.05 or less grains per cu ft, given a coal size distribution shown in the basic data in Table I. It is noted that the dust laden coal stream has a size distribution by weight of 33% in the 2 to 5-p size range and 55% in the 0 to 2-p size range. The particle size distribution given in the specifications does not detail the percent by weight less than 1-p in size, and more important, the size distri-

Jan 1, 1963

By R. Schuhmann Jr., P. J. Ensio

As a first step in a study of the physical chemistry of copper-smelting slags, experimental measurements were made of the oxygen pressure of simple iron-silicate slags in equilibrium with solid iron. The experiments consisted in bubbling CO2-CO mixtures through the slags in iron crucibles and in finding the equilibrium ratios of CO2 to CO. From the data, activities and partial molal heats of solution FeO and SiO2 in the slags were calculated. THE ternary system FeO-Fe,O,-SiO, is the simplest system which can represent adequately the important chemical properties of the slags formed in matte smelting and converting. Accordingly, the experimental study of slags in this system was started as the first step in a general research program on the thermodynamics of copper smelting. A recent paper by one of the authors' outlines this general research program and shows in some detail how information on thermodynamic properties of iron-silicate slags, especially information on oxygen activities, is essential to a full understanding of the chemistry of copper smelting. The FeO-Fe,O,-SiO, system also is the basis of slags produced in other processes, such as acid steelmak-ing processes, and is important in relation to the behavior of silicious refractories. In addition, data on the physicochemical properties of iron-silicate slags should contribute to a general understanding of the nature of slags, since at present such data for all kinds of slags are relatively meager. The principal quantitative information already available on iron-silicate slags consists in Bowen and Schairer's constitution diagram for slags in equilibrium with metallic iron,' Darken and Gurry's complete and precise thermodynamic study of the Fe-0 system,",' and Darken's study of the phase equilibria in the Fe-Si-o system. From the results of this previous work, it was clear that a complete thermodynamic study of ternary iron-silicate melts was a major undertaking which would have to be broken down into parts, including the working out of the ternary equilibrium diagram as well as extensive activity measurements on slags. The experimental measurement of oxygen activity, using CO-CO, and other gas mixtures as yardsticks of oxygen activity, was the basis of Darken and Gurry's study of the Fe-0 system. From the experimental measurements of oxygen activities over a large range of temperatures and compositions, they were able to calculate the activities of other species such as Fe, FeO, Fe,O,, etc., and also the other important thermodynamic properties in the Fe-0 system. At the start of the present investigation, it appeared that the same approach would be fruitful for the study of iron-silicate melts. Also, the experimental measurement of oxygen activities seemed particularly attractive because oxygen activity itself is a quantitative measure of oxidizing or reducing power, which is an important variable in copper-smelting slags. The slag compositions in equilibrium with yFe appeared most suitable for the first investigation of this kind, for the following reasons: 1—Good constitution data2 were available. 2—An iron crucible could be used, thus affording a simple solution to the refractory problem without introducing contarnination. 3—With Fe activity fixed, at unity with respect to a standard state of solid Fe, measurements of 0 activity could be followed by simple calculations of activities of FeO and other Fe-0 complexes and of SiO,. 4—This starting point afforded several possibilities for cross-checking with previous work,

Jan 1, 1952

By H. H. Rachford, D. W. Peaceman

A system of partial differential equations describing miscible displacement of fluids in porous media is derived The system takes into account the influence of gravity, the spatial distribution of permeability, diffusion, and fluid viscosities and densities. A numerical procedure for approximating solutions to the differential systems has been tested for a horizontal two-dimensional geometry. In end-to-end displacements of oil with less-viscous solvents, the numerical solutions exhibited fingering qualitatively similar to that observed in laboratory models. Small random spatial variations in permeability about the mean value are sufficient to initiate fingering. Quantitative comparisons of computed results with laboratory data show good agreement. INTRODUCTION Miscible flooding of oil by solvent is receiving increasing consideration for field use.1-4 Unfortunately, such floods potentially present severe problems in loss of recovery through the by-passing of oil by expensive solvent. Consequently, their economic evaluation requires sound techniques for predicting recovery. The purposes of this work are to present a finite-difference method for calculating the multidimensional displacement of oil by solvent and to investigate the validity of the method by comparing results of calculations with data from displacements in laboratory models.5 The formulation of the method to simulate the model experiments treats a case of limited scope in the description of solvent flooding in the reservoir. The model experiments were carried out with fluids that were assumed to form an incompressible, ideal, two-component system with constant diffusivity. Establishing the validity of the method even in its present form provides a major step toward the goal of quantitatively evaluating individual solvent flooding projects. First, it demonstrates the fea- sibility of calculating the course of displacements which are dominated by an inherent macroscopic instability, i.e., viscous fingering. In addition, the method provides almost the only practical means of examining the effect of the size and extent of reservoir inhomogeneities on the development and propagation of the fingers, taking into account the important influences of diffusion and gravitational segregation. It might be inferred that such studies cannot yield results of practical value because in displacements dominated by macroscopic instability the pattern of finger development and the resulting performance should depend critically on small variations in the geometric distribution of reservoir inhomogeneities. The present method offers the capability inherent in computational techniques of predicting performance reproducibly with any arbitrary distribution of reservoir properties and, as such, provides a means of evaluating the sensitivity of behavior to uncertainties in reservoir definition and, thus, assessing the reliability of prediction of performance. THE PHYSICAL PROBLEM DATA USED FOR QUANTITATIVE TESTS OF CALCULATIONS In this work, a calculation for treating the miscible-displacement process is tested by comparing calculated results with experimental data. The experiments chosen for comparison have been described in detail,5 and the quantitative features are summarized in a later section. Briefly, oil was flooded by solvent of equal density from a thin rectangular channel in Lucite packed with uniform Ottawa sand. Under these conditions, two-dimensional geometry is considered adequate to represent the process. Inasmuch as the present work is concerned with testing of the method by comparison with specific two-dimensional data, the system of equations presented in the succeeding section will be oriented to defining the physical system specifically. THE DIFFERENTIAL EQUATIONS Suppose x, y to be a Cartesian co-ordinate system and define h(x, y) to be the height of a point above a horizontal reference plane. It will be assumed

By R. H. Hiltz

Three titanium alloys, known to provide a mechanically unstable p structure after quenching, were selected as material for a study of the Origin and nature of stress-induced transformation. Data from hardness and tensile tests were used to study the initiation of marten site and the effect of the transfornation and transformation products on the mechanical properties of B titanium, particularly the stress-strain relationships. With the exception that the materials investigated showed a natural aging reaction in the P condition, their behavior is analogous to that of' stress-transforming stainless steels. The transformation starts after initial yielding and is essentially complete just prior to .maximum loading. Once initiated, the martensite itself is capable of propagating the transformation. The deforn-ing materials show low strength, high ductility, and a high strain-hardening exponent. In metallic systems which undergo a diffusionless transformation on cooling, it is possible to produce a similar transformatiorl by plastic deformation under special conditions. Dewez' in delineating Ms temperatures for the titanium-rich end of most titanium binary systems:, established the point of intersection of Ms with room temperature for those particular alloys. Alloy compositions which would quench to mechanically unstable p were established from these data. The first investigation into the possibility of stress-induced transformation occurring in titanium met with little success. It was concluded that such a transformation in titanium was possible only under very special conditions. Continued effort in this direction, however, showed certain Ti-Mo binaries to be highly susceptible to this phenomenon. This fact was utilized by Liu to study the habit planes of martensite and by Machlin and wenig5 to study the effects of stress and plastic deformation on the P to a prime transformation in titanium alloys. The first investigation of the mechanical properties of these alloys was conducted by Holden and his coworkers at Battelle Memorial Institute.' The potentiality of the alloys i3s sheet-forming material was recognized, and the data reappeared in a survey of this type of titanium material.? The initial optimism was tempered by the difficulties in obtaining homogeneous 2i-Mo alloys in commercial-size ingots. The increasing need for good sheet-forming material kept investigation in this direction active, resulting in the development of a commercial alloy of this type.' This study was undertaken to obtain a clearer picture of the mechanism of stress-induced transformation in titanium. MATERIALS AND TEST PROCEDURE Three alloys were selected for this investigation: one a Ti-Mo alloy; another a Ti-Cr-Mo alloy, known to deform martensitically; and the third, a Ti-V alloy with an Ms below room temperature and an Md above room temperature, but known to contain w phase in the quenched condition. The compositions of these alloys are given in Table I. These alloys were cast as 5-lb experimental ingots, employing skull melting. The ingots were hot-forged into 1-in.-round bars. Two 72-in.-thick slices were cut from each bar for hardness testing; the remainder was cut into 0.252-in. tensile blanks. The tensile blanks and one hardness disc from each bar were heated in salt to a temperature 100°F above the 0 transus, held 2/2 hr, and brine quenched. X-ray analysis showed the pieces to be 100 pct 8.* *Omega cannot be distinguished from p by the technique employed. Tensile blanks were machined to size after this treatment. TO prevent transformation due to machining stresses, the shoulders and gage length were ground. True stress-true strain curves were

Jan 1, 1960

By H. A. Lequear, J. D. Lubahn

A sudden change from one constant strain rate to another during a tensile test causes an unusual transient in aluminum alloy 61ST. A sudden change from one constant stress to another during a creep test does not produce an unusual transient in this material. A specially devised test shows that the nature of the previous steady-state conditions, which the sudden change disturbs, is one of the factors affecting the transient behavior. Some of the observations may be explained in terms of strain aging. USUALLY a sudden change during a tensile test from one constant strain rate to another causes a small, abrupt vertical shift from one stress-strain curve to another, parallel one. This behavior is apparently characteristic of all metals at very low temperatures and of many metals for wide ranges of temperature. It is illustrated in Fig. 1 for some early data on 61ST tested at — 195oC, and in Fig. 2 for some recent data on copper at room temperature. The greater accuracy of the results of Fig. 2 over previous ones' is due to the equipment shown in Fig. 3. Three levers apply the bulk of the load, while the changes in the small remainder load are indicated sensitively by a low capacity load scale. Refs. 2 and 3 describe the recording equipment. A test like those in Figs. 1 and 2 provides the information necessary to determine the rate sensitizrity, n,4 of the metal. ( a log s ) ? log s n = ------------------------ ----------------------------= ( a logs e, ? log e logS2-logS1 log S2/S1 log i2 — log e1 log e2/e1 where S equals stress: e is logarithmic strain or natural logarithm of (final gage length/initial gage length);" i is strain rate or de/dt: and n is rate sensitivity, rate of change of log stress with log strain rate at a given strain, or (a log S/a log i),. Normally, the rate sensitivity, n, is positive, which means that an increase in stress causes an increase in rate. However, certain exceptions, attributed to strain aging,5 have been observed. Fig. 2 indicates how rate sensitivity values are derived from the measurements. Table I summarizes the rate sensitivity values derivable from Fig. 1. In contrast to the behavior of copper in Fig. 2, the aluminum alloy 61ST, which strain ages when tested near room temperature, exhibits an unusual transient when the strain rate is suddenly changed,' as shown in Fig. 4. Initially, there is a comparatively sudden vertical shift of the stress-strain curve in a direction corresponding to the rate change, but then the curve gradually shifts back again. The initial shift is not as sudden as might be expected, because the testing machine is unable to achieve the impressed rate change instantaneously, as illustrated in Fig. 2. The behavior shown in Fig. 4 will be referred to as a tensile transient, the word tensile implying that the behavior was observed in a tension type of test. The corresponding test, where the strain rate is maintained constant except for the sudden change, will be called a tensile transient test. A creep transient, by contrast, will be defined as a temporarily abnormal behavior in a creep transient test, where there is a sudden change from one constant load to another.' In Fig. 4, a naive look at the gradually decreasing and then increasing load after a sudden rate incre-

Jan 1, 1957

By M. H. Childs, K. E. Merklin

Blast furnace feed specifications are being revolutionized because of the large increase in iron pellet production from low-grade ores of the Lake Superior region. Recent laboratory findings aimed at a better understanding of factors influencing the physical quality of pellets produced from taconite concentrates are reviewed. Tests for determining better product quality, the effects of additives, and production factors are discussed. The current revolution in specifications for blast furnace feed is typified to a great extent by the recent tremendous growth in iron ore pellet production from low grade ores of the Lake Superior region. Indeed, pellets are responsible, at least in part, for many of the stringent new requirements that are being placed on the suppliers of furnace raw materials. As the tonnage of pellets going to the steel mills increases, more and more attention is being turned to the detailed characteristics that make this product an ideal blast furnace feed. Control of quality in pellets is similar to that of any other product. A basic understanding of the factors that are encountered in their production, from the crude ore through to the finished product, is required, along with careful process supervision and frequent quality checks. Normally, chemical grade is not a problem. This usually can be controlled within the desired specifications by minor operating variations in the con- centrator. While future changes in requirements may result in the need for self-fluxing pellets1 or other chemical changes, today's quality demands center primarily around the physical characteristics of the pellets. It is in this area that many of the recent laboratory studies have proven helpful to the operating plants. A better understanding of the physical factors involved in pelletizing has provided the basis for a continuing program of quality improvement. Tigerschiold and Ilmoni2 and Cooke with various associates are responsible for the development of much of the basic understanding of the pelletiz-ing process. These earlier findings have served as a most useful background for this work. LABORATORY PROCEDURES Magnetite concentrates produced from eastern Mesabi Range magnetic taconites were used for all of the work reported herein. The sizing and chemical analysis of a random sample are shown in Table t Additives usually are mixed with the concentrates in taconite plants before pelletizing to improve the strength characteristics of the unfired balls. Ben-tonite, soda ash, and coal are in common use. The bentonite used in these studies was of the Wyoming sodium type; the soda ash was a standard commercia1 brand. Anthrafines were used where coal was called for. The procedure for making batches of balls at the Hibbing Laboratory is as follows: In an especially designed mulling equipment 2800 g of dry ore are mixed for 10 min with the desired additives. Water is added and mixing continued for 1 1/2 min. The charge is removed from the muller and put through a mechanical shredder to break down any lumps that may have formed. sixty previously formed seed balls are placed in the balling drum, which is in fact a 6 x 16-in. diam airplane tire turning at 53 Wm. Two thousand grams of the shredded concentrate and additive mix are fed by hand into the tire at a con-

Jan 1, 1961