Search Documents

By E. J. Onuscheck

With passage of the Federal coal' Mine Health and Safety Act of 1969, many challenges have been directed to the industry's mining and safety engineers. The most critical challenge is that of ventilation. The language concerning ventilation in many sections of the Act could very well deprive or limit the mining engineer in planning the ventilation for new mines, since the Act requires federal approval of all ventilation plans. Dependent upon approval and of course application of the law, this could, due to economics, prohibit the opening of new mines in some cases. Petitions for modification of some mandatory standards relating to ventilation have already been filed under Section 301(c). It is quite possible that many of these standards may be modified.

Jan 1, 1970

By Greenleaf W. Pickard

The Parral mining district is located in the southern part of the state of Chihuahua in north-central Mexico, in the border area of the physiographic provinces of the Mesa Central and the Sierra Madre Occidental, at an average elevation of 1700 meters. This is one of the most highly productive lead-zinc districts of Mexico. The principal operating units are the Prieta, Santa Barbara and San Francisco mines. The Santa Barbara-San Francisco del Oro area is located 25 kilometers to the southwest of the Prieta mine. It has a system of generally narrow veins in fault zones, entirely in the shale series. The Santa Barbara veins follow long, persistent fractures, while the San Francisco del Oro veins are in relatively short fractures. The Veta Colorada structure, located 5 kilometers to the northwest of the Prieta mine, was a major producer of high grade silver ore with low lead-zinc content during the eary 1900's. It follows a major north-south fault zone in andesite and is noteworthy for a number of wide oreshoots of limited vertical extension along its 10 kilometer length. The Palmilla and Esmeralda mines are on the south end of the Veta Colorada structure. Numerous smaller mines are operating in the Parral area. During the last 45 years the Prieta mine has produced approximately 15,000,000 tons of complex sulfide ore, principally from one single vein. Extensive stoping has

Jan 1, 1970

By C. Laird, C. E. Feltner

IN previous workif2 we have shown that the steady state cyclic stress response of pure copper tested in the low cycle fatigue range was determined only by the plastic strain amplitude and the test temperature. hus, copper, at any given temperature, has a unique cyclic stress-strain curve independent of prior history. Cu-7.5 wt pct Al, however, showed a stress response for a given plastic strain amplitude and temperature which depended sensitively on prior history. From this work we predicted that all wavy slip mode materials, typified by copper, would have a unique cyclic stress-strain curve; whereas all planar slip mode materials, typified by Cu-7.5 pct Al, would not. This prediction has now been tested for a number of materials of widely varying crystal structure and slip character. Wavy slip mode materials studied include: Fe. Al, Cd, disordered FeCo-V, and Cu-3 pct Co (quenched and aged). Planar slip mode materials investigated include: Fe-3 pct Si, ordered FeCo-V, disordered Ni3Mn, ordered Ni3Mn, and Zn. The treatment and properties of these materials are listed in Table I. Cyclic stress-strain curves were determined by controlled plastic strain tests (see Ref. 1 for details) at room temperature for all materials in both the initially annealed and cold worked conditions. The results obtained for iron and Fe-3 pct Si are typical of wavy and planar slip mode materials, respectively. The cyclic stress-strain curves for these materials are shown in Fig. 1 and it can be seen that iron has a unique cyclic stress strain curve while Fe-3 pct Si does not. Aluminum, Cu-3 pct Co, and disordered FeCo-V showed a behavior like that for iron, while Ni3Mn in both the disordered and ordered condition behaved in the same manner as Fe-3 pct Si. These results therefore support the general conclusion drawn in Refs. 1 and 2 that. for wavy slip mode materials, a cyclic mechanical equation of slate it =t.(hcp, i. T) is approximated in which the satura-

Jan 1, 1970

By B. Barlin

Introduction The development of metallurgical practice at lambia Broken Hill is directly related to the change in mineralization of the deposits as mining progressed from the open pits to underground. The ore consists of a mixture of lead and zinc sulphide minerals, together with abundant silicates, carbonates and other oxidized minerals of these metals. The irregularity that characterizes the shape of the orebodies applies equally to their mineral content, which makes ore reserve calculations and the maintenance of a balanced feed to the metallurgical plants, problems of some magnitude. History The Rhodesia Broken Hill Development ,Company was formed in 1904 after an Australian prospector, T. G. Davey, discovered the outcrops in 1902. In the early years zinc ores were calcined for shipment but this proved uneconomic. Cerussite outcrops were then smelted directly in blast furnaces for the production of lead until 1929 when ore above the water level was exhausted.

Jan 1, 1970

By Austin Park

More metal at a lower profit continued to be demanded from the mining industry during 1969. Correspondence with U.S. and Canadian mining districts and a study of the mining literature indicate that, a large majority of the mines are keeping pace and rising to the challenge. Continued research and better understanding of pneumatically applied concrete for roof and wall support was of top interest. Many mines are changing mining methods for the adaptation of load-haul-dump units. The most recent adaptation being the use of these units in cut- and-fill stoping and in various types of development headings. Boring methods and use of two- or three- boom drilling jumbos mounted on rubber tires for versatility continues to be the favorite approach to the demand for rapid excavation.

Jan 1, 1970

By R. Maxwell

"At Opemiska Copper Mines we have, over the past four years, operated 300 cu. ft. and 700 cu. ft. flotation cells as roughers, scavengers and cleaners and found them competitive with conventional flotation cells on a cu. ft. basis on all phases of flotation.The economics are very much in favour of the large cells for the following reasons:Power consumption is 35 to 40%, of conventional machinesMaintenance cost is about 20 to 25% of conventional machinesSpace required is about 30 to 35% of conventional machinesInstallect cost is about 40%, of conventional machinesTo put the above in more concrete terms below is what we are doing at Opemiska:1. Increasing mill capacity from 2000 to 3000 T.P.D. and will use only 5Cff, of the flotation space for 3000 tons that we used for 2000 T.P.D. with conventional flotation machines.2. Will require only 10 flotation cells for 3000 tons against 85 conventional cells for 2000 tons.3. Cells will be, one 2000 cu. ft. rougher, six 700 cu. ft. scavengers (2 banks of 3 each) and one cleaning circuit using three 300 cu. ft. cells.You will note that in this 3000 ton per day plant, we are going to have only ten cells. Had we put in conventional cells for this tonnage, we would have needed 120 cells."

Jan 1, 1970

This report summarizes a study conducted by The Service Bureau Corporation for the Bureau of Mines. Evaluations reflected herein are the result of an independent investigation by The Service Bureau Corporation of methane considerations during coal mining. The study is oriented toward a systematic analysis of methane control problems during coal mining. Throughout this study an attempt has been made to ensure considerations consistent with the recently enacted Federal Coal Mine Health and Safety Act of 1969. References to reports by The Bureau of Mines and others are made for the purposes of establishing identification between the evaluations contained in this report and on-going research efforts.

Jan 1, 1970

By J. R. Johanson, A. W. Jenike

"A solid flows out of a bin provided that the outlet is sufficiently large so that stable obstructions to flow do not develop. The geometry and surface finish of a bin and the construction of the feeder determine the type of flow pattern which then develops within the bin. The pattern, in turn, affects the uniformity and consistency of the stream, segregation, caking, degradation, live capacity and the action of level measuring devices.In most bins now in operation, material flows toward the feeder through a channel which forms within the stationary solid. This is referred to as funnel flow. If the hopper of the bin is sufficiently steep and smooth and the feeder capable of drawing material across the whole area of the outlet, all the solid flows whenever the feeder is in operation; this is referred to as mass flow. Mass-flow bins are usually far superior to funnel-flow bins; their initial cost is also usually higher.The paper describes the design criteria used in deter-mining the minimum bin outlet needed for flow and the bin geometry needed for mass flow."

Jan 1, 1970

By Jack Fox

It is some time since these columns have contained a report to the members on just what is doing in the Society of Mining Engineers. Accordingly, even though it is now a month and a half after the Annual Meeting in Denver, it seems that a report on what transpired at the SME Board Meeting there is in order. First off, I'd like to say that the Members of the Society of Mining Engineers can be justly proud of their representatives on the SME Board of Directors. Of 21 members and officers on the "Old" Board of Directors, 19 were present. President Brower Dellinger chaired the "Old" Board Meeting. Of the 21 members on the "New" Board of Directors, every single one was present. James Boyd, the outgoing AIME President, attended and expressed his appreciation of the support that the Society of Mining Engineers had given him during his year as President of AIME. Members Multiply Robert S. Shoemaker, Chairman of the SME Membership Committee, pointed out that based on figures available, the Society of Mining Engineers had a net gain of 523 members during 1969 while SPE had a 490 net gain and The Metallurgical Society a net loss of 159. This is probably the first time in the last 20 years that anyone has exceeded the SPE net gain in membership. While SME still has a way to go to surpass the Society of Petroleum Engineers in number of members, we did have, on December 1, 1969, 16,369 members. However, 419 were dropped at the end of that month leaving us with a net for the year of 15,950.

Jan 1, 1970

By R. Venkataramani

Even though the mining industry in 1966 produced more ore than ever before, current extraction rates are only a fraction of what is expected in the later years of the twentieth century. Nearly 90% of all metals and-mineral products consumed last year was recovered by open-pit mining. This has placed great pressure on this segment of the industry which has, consequently, resulted in some spectacular developments. With increasing size of projects, the need for increased sophistication of engineering, planning, management and administration of modern mining installations has never become more apparent. The design of complete systems for the mine and plant that fit the mould of today's business and social environments is undergoing an evolutionary process. Traditional concepts in mine development and operations are being sidestepped in favor of new ideas and principles. As the overburden thickness increases, materials handling presents a major problem to mining companies, especially those concerned with the mass production of ore and waste from low-grade deposits. The profit margin here is likely to be significantly less as to take chances with capital investment. 1 Constant efforts are needed to improve upon productivity if the ore is to be economically mined.

Jan 1, 1970

By A. U. Seybolt, J. L. Walter

The effects of the addition of 0.5 pct ThO2 (by volume) on the grain structure and texture of heavily rolled tungsten sheet were studied. Powder metallurgy ingots were rolled to sheet, 0.001 in. thick, and samples were annealed at temperatures from 1200" to 2500°C. The heating rate to the anneal temperature was varied from 3°C per min to 6500°C per min. The orientations of the largest grains were determined by X-ray diffraction. The disposition and loss of the ThO2 particles were determined by electron transmission microscopy and the activation energy for remozlal of the thorium was measured by mass spec-troscopy. Heating rates below about 800 C per min produced large, highly elongated grains. Higher heating rates produced much smaller, less elongated grains. The grain structure and texture of the slowly heated samples were identical to the structures and textures of heavily rolled tungsten foil conventionally doped with aluminum, silicon, and potassium but the recrystallization temperature was about 200°C lower. There was stringer-like dispersion of larger ThO2, particles superimposed on a uniform dispersion of fine particles. The particles increased in size and decreased in number with time at temperature. The loss of particles was accompanied by formation of grain boundary voids and surface pits. The activation energy for removal of the ThO2 was about 40 kcal per mole. THE addition of small amounts of aluminum, silicon, and potassium to tungsten has a profound effect on the grain structure of heavily rolled sheet1 or drawn wire."4 Thus, "doped" tungsten sheet, rolled to 0.001 in. and recrystallized, develops a structure consisting of large, highly elongated grains of a specific preferred orientation.' It has been shown that the elongated grain structure was caused by a dispersion of particles, identified as mullite (Al6si2O13),1,4 in strings parallel to the rolling direction in sheet or to the wire axis. Identical results were obtained for iron which was doped with glass and rolled to 99 pct reduction of thickness and recrystallized.5 Again, the glass particles were disposed in strings parallel to the rolling direction. The strings retarded boundary migration in the lateral direction but grain growth was able to proceed relatively unhindered along the direction parallel to the strings.1,5 In both the doped tungsten and the glass-doped iron, the particles were highly plastic during the high-temperature deformation. Since mullite melts at about 1800°C and the glass employed melts at about 810ºC, the particles were probably molten during the recrystallization anneals (above 1800°C for the doped tungsten and 850°C for the glass-doped iron). There are indications in the literature that particles not plastic at hot-working temperatures (up to about 1550°C for tungsten) may also impart elongated grain structures by primary recrystallization of worked alloys. Examples are dispersions of tantalum carbide in chromium,6 and ThO2 in nickel where elongation ratios of recrystallization grains of up to 20:l were obtained.7 Strings of ThO2 particles in W-ThO2 wire have been seen by Rieck.8 In view of the available evidence, it was of interest to extend the previous studies of doped tungsten and glass-doped iron to include W-ThO2 sheet materials. Of specific interest was the grain structure, preferred orientation, recrystallization temperature, and disposition of the thoria particles. EXPERIMENTAL PROCEDURE A) Material Preparation. About 380 g of blue tungsten oxide (approximate composition WO3) was wet with a sufficient amount of aqueous solution of Th(N03)4 to provide 0.5 vol pct ThO2. The wet mixture was stirred to distribute the doping material uniformly throughout the powder. Stirring was continued while the mixture was heated until it was completely dry. The dried, doped mixture was then reduced to tungsten powder by heating at 850°C for 2 hr in a stream of dry hydrogen. The doped tungsten powder was hydropressed in a rubber sack at pressures of 50,000 psi to produce an ingot approximately 2 cm diam by about 5 cm long. Next, the pressed cylinder was heated in hydrogen for 2 hr at 1350°C to render it strong enough for machining. The cylinder was machined to a square cross section of 1.6 by 1.6 cm with rounded corners. Finally, the bar was heated for 2 hr at 2150°C in hydrogen to provide a sufficiently dense mate rial to withstand the hot-rolling operation. Prior to rolling, the ingot was reduced in thickness by about 20 pct by forging at 1550°C. This served to increase the density of the material and to widen it slightly for better rolling characteristics. The forged ingot was then rolled at 1550°C to a reduction of 45 pct of thickness, following which the rolling temperature was reduced to 1300°C and the slab rolled, with several reheats, to 0.050 in. Further reduction was accomplished at 800°C in tungsten sheet packs to a total reduction of about 99.85 pct. B) Annealing Treatment and Observation of Structure. Small samples of the rolled foil were annealed in one of two ways. Some of the samples were suspended in a covered tungsten can and annealed at temperatures from 1200" to 2200°C for various times in vacuum of l0-7 to 10-6 mm Hg. The can was heated by induction and temperatures were determined by optical pyrometry

Jan 1, 1970

By A. D. Taylor

In recent years there has been an increasing effort to attract students into engineering. Obviously, the effort is necessary because young people find the image of the engineer unattractive. Some of the things they find unattractive are real, some are not, and part of the problem is too little knowledge of the field. First, I would like to comment on some of the ideas which appear to be justified. Engineering is a difficult course of study. True, and nothing can or should be done to change this. The education is continually being improved, but it is unlikely to be made easier.

Jan 1, 1970

By Donald R. Corson

The effects of adding varying amounts of type I portland cement to hydraulic backfill are described. Six different fill materials, each with three cement contents, were tested to determine which fill properties have the most significant effect on cemented strength. Two strength prediction equations, based on the physical and mineralogical properties of the fills, were formulated, and these accounted for 96 and 99 percent of the variability of the data. Results of both unconfined compression and triaxial tests of cemented fills are given, and the effects of moist curing and fill densification on the 7-day strength are discussed. Permeability test data are given, and the decreased permeability due to cement content and curing time is shown graphically. A discussion of the limited evaluation of other potential stabilization additives (expansive cement, fly ash, sodium silicate, gypsum, and asphalt) is included in an appendix.

Jan 1, 1970

By Herbert Dumont

Since 1947 Preussag AG Metal1 has used a process for fuming zinc- and lead-containing dump slags and residues from vertical retorts in a half-shaft furnace at their Hüttenwerk Harz at Oker. Contrary to the classic shaft furnace, only one side of the half-shaft furnace is fitted with tuyeres. The opposite side is a bricked slope. The upper shaft section is piped. There zinc vapor and combustion gases are burned by sucked-in air. Gases and fume enter a boiler located on the rear length of the furnace via a piped fire bridge. The gases are then cooled in an air preheater prior to dedusting in a bag filter. The furnace is charged with briquettes of metal-bearing materials, coal and pitch. The leaded zinc oxide fume is comparatively rich in both lead and zinc. The copper in the feed is recovered in low-copper matte. The slag can be discarded. The Oker plant comprises 13 furnaces of 2.9 to 4.9 m length in the tuyere area. Two other plants in Italy and Spain are processing zinc leach residues by this technique. Fuming of zinc-lead-oxide ores as well as partially roasted sulfide- zinc-lead-copper bulk concentrates will be possible.

Jan 1, 1970

By J. M. Singer

Approximate equations were used to compute the radiant heating of float coal dust by approaching flames for various combinations of particle size, dust path length (flame travel distance), flame temperature, dust concentration, flame speed, and flame length. The computed values of degree of radiant heating and flame travel distance for preheating the coal dust should be useful for estimating the increased fire hazard of different coal dust-flame parameters. Some findings relating to coal mine disaster control are-- 1. Coal dust clouds cannot be preheated to ignition temperatures by flame radiation alone when deflagration speeds are greater than about 10 m/sec. Thus it appears unlikely that radiation alone will cause flames propagating in dust clouds to accelerate to detonation velocities. 2. Small particles in the radiation flame zone caused by burnout and soot formation increase the degree of preheat significantly. 3. The concentration of float coal dust (radii = 3 to 40 microns) for maximum radiant heating usually peaks in the range 50 to 200 mg/liter of air.

Jan 1, 1970

Copper ores are widespread through the Lachlan Geosyncline in New South Wales, but the size of the orebodies is generally small. The largest deposits occur in the Cobar district where an iron-rich copper ore type has been developed in Silurian sediments. Similar de- . posits to the south in the Nymagee-Mt. Hope-Melrose districts are related to quartz feldspar porphyries of probable Upper Silurian age. Deposits to the east in the Tottenham-Girilambone districts are less complex mineralogically and are possibly related to an ultra-basic source.Copper mineralization in the Highlands further to the east is often developed in Ordovician andesite or MiddleUpper Silurian acid volcanic sequences. The deposits in andesite contain considerably less galena than those in the acid extrusives; a small proportion of the mineralization in the latter has texture, mineralogy, and geologic setting analogous to the Captain's Flat copper-Iead-zinc deposit, for which both syngenetic and epigenetic origins have been proposed. Very little copper mineralization occurs within granitic material, even though rocks of this type ranging from Silurian to Carboniferous in age crop out through much of the geosyncline, and the extent of the metal contribution from plutons to the deposits in volcanics is open to question. Numerous deposits are known in the contact zone around Mid-Devonian and Mid-Carboniferous granites, however, and the formation of these orebodies has undoubtedly been influenced by plutonic activity. Small occurrences of copper with associated minor nickel sulphides are developed in a narrow serpentinite belt of possible Lower-Middle Devonian age in the southern section of the Highlands.

Jan 1, 1970

By Swan. J. D.

Repairing and replacing equipment damaged by corrosion and abrasion is a major cost throughout industry, especially in the coal mining industry. Careful study of corrosion and abrasion, and their causes and effects, often leads to the selection of alternative materials which can provide economies. What, actually, are corrosion and abrasion and how do they contribute to the cost of operating a coal-preparation plant? Furthermore, what can be done to alleviate corrosion and abrasion problems and thereby reduce operating costs? These questions will be answered in the following discussion, which will describe the destructive action of corrosion and abrasion and give several methods for controlling corrosion and abrasion losses, These methods include the substitution of stainless steel for carbon steel, the use of corrosion inhibitors, and the use of design to eliminate corrosion and abrasion. Examples will be given illustrating the benefits that have resulted in plant tests when these methods of controlling corrosion and abrasion were used.

Jan 1, 1970

By Paul H. Johnson

A field investigation was undertaken to study some of the. chemical and hydrological aspects of leaching an impermeable copper sulfide waste dump. Samples, obtained by means of a Becker pneumatic drill from Kennecott's Chino "J" dump, near Silver City, New Mexico, were analyzed for their moisture and soluble-salt content, Fe +2 , Fe +3, Cu +2 , and H in their interstitial solutions and non-water - soluble Fe, Cu, and S in their rinsed residues. Conclusions were drawn from these data concerning the overall chemistry of impermeable-dump leaching and the hydrological characteristics of the dump.

Jan 1, 1970

By P. Shahinian, H. H. Smith, M. R. Achter

Crack growth rates are measured at elevated temperature in a resonant fatigue machine from vibration frequency decreases calibrated in terms of crack depth. Crack growth rates in Type 316 stainless steel at 500º and 800°C show a sharp increase with oxygen pressure in an intermediate pressure range and little or no change at high and low pressures. At 500°c, I torr of oxygen reduces the fatigue life by almost a factor of 100 in comparison to that in vacuum and raises the growth rate of shallow cracks by the same At At 800°C the effects are smaller. Changes in slope in the crack growth rate curves are discussed in terms of a model in which rates of surface production and of surface coverage by gas are compared. The use of a calculation method in which the surface exposure time is equal to X/v, where x is the interatomic spacing and v is the growth rate, makes it possible to obtain order of magnitude agreement at 500°C between the observed pressure and the predicted pressure at these slope changes. At 800°C oxidation becomes a .factor and the data cannot be treated by simple adsorption theory. THE decrease in the fatigue life of metals as a function of gas pressure usually follows a stepped curve with virtually all of the decrease concentrated in a sharp drop in a transition zone at intermediate pressures and little or no change at low and high ranges. A number of models, differing in the details of the mechanism, have been offered to explain the shape of the curve. Measurements of crack growth in aluminum as a function of gas pressure by Bradshaw and Wheeler' and Hordon2 demonstrated opportunities for quantitative comparison to evaluate the proposed models. Since comparable data were lacking at high temperatures, in the present work rates of crack propagation were measured in Type 316 stainless steel at 500" and 800°C as a function of oxygen pressure. Choice of this material was dictated by two considerations; it is stiff enough at these elevated temperatures to resonate with the regenerative drive on our fatigue machine and it is known to display a large effect of environment. A new method of calculation is described to predict the gas pressure at the critical point. EXPERIMENTAL PROCEDURE Because of the difficulty of measuring crack depths directly at high temperatures, an indirect method was developed based on the decrease in the resonant frequency with the growth of a crack. A reversed bending, constant amplitude fatigue machine, described previously,3 vibrates a specimen at its resonant frequency, automatically records any changes in it and shuts itself off after it has reached a preset value of frequency decrease. The record of frequency change is used to determine the rate of crack growth. Sheet type specimens of Type 316 stainless steel, Fig. 1, incorporated a sharp, shallow notch to localize the formation of a single crack. After machining, they were annealed in a vacuum of l0-6 torr either at 1066" (lot A) or 871°C (lot B) and then electropolished in an acetic-chromic acid solution. Bending strains were measured at 500" and 800°C by an optical technique4 and reported as total strain without correction for the notch. At 500°C, the 0.141 pct strain was 0.085 pct elastic and the remainder plastic. At 800°C the 0.062 pct strain was all elastic. To convert frequency decrease to crack length, calibration curves were obtained by interrupting the vibration at stated intervals of frequency decrease. The crack depth was measured microscopically at a magnification of X400 and reported as the average of the measurement on each edge. Some specimens were sectioned for crack measurement while others were returned to the machine and fatigued further. There was good agreement between the two methods. Before beginning the vibration, the vacuum chamber was first evacuated cold to 1 x 1O-6 torr, then heated to the operating temperature and held there until the pressure was again reduced to 1 x10-6 torr at which time oxygen was introduced to the desired pressure. In this investigation the vibration frequency was nominally 10 cps and a decrease of 0.6 cps was taken as the failure point. The choice of the frequency decrease to represent failure has no appreciable effect on the fatigue life because the crack is growing very fast at this point.

Jan 1, 1970

By Pradeep K. Rohatgi, David N. French, Surendar M. Jain, Clyde M. Adams

The effect of solute concentration and diffussivity on dendritic solidification of dilute binary aqueous solutions has been investigated; chlorides of sodium, potassium, lithiurn, and hydrogen were used as solutes. Dendrite spacings were measured in spherical droplets and in unidirectionally solidified specimens. In the droplets, the freezing rate is uniform over the entire specimen and remains constant from the beginning to the end of solidification. In the unidirec-tionally frozen specimens the freezing rate varies within a given volume of liquid during the solidification cycle; in addition, it varies with location uithin the specimen. Mass transport analysis has been used to calculate the supersaturation, AC, supercooling, AT, and transcerse growth velocity, V , of ice dendrites during solidijication from the dendrite spacings measured in the droplels. The ice dendrite spacing in a given solute increases lineary with the solule concentralion at a fixed freezing rate. The ualues of AC, AT, and V increase with an increase in solule concentration. The value of AT in the aqueous solution is of the same order of magnitude as in Al-Cu alloys of the same concentration; the dendrite spacings in Al-Cu alloys also increase with an increase in solute concentralion. For the jamily of aqueozrs solulions incestigaled, the dendrite spacing increases linearly with solute diffusivity at fixed values of concertralion and freezing ,rate. The values of ?C and ?T decrease, whereas the value of V increases with an increase in solule diffusicity. The measured dendrite spacings , L, can be expressed by the relatiun: L = K;(dfa/db) 4 K2C +K3D - K4 where dfs/d? is-change in fraction solid with time (freezing rate), C is solute concentration, D is solute diffusivit K1 = (14 * 4)pmin1'2, K2 = (70 ± 10) pl (moles)- K3 = (24 r 3) . l05 psec (cm)-~. K. = -132 ± 10) 1. FREEZING of solutions is almost always an ehunmix-ing" process because the composition of crystallizing solid generally differs from that of the parent liquid. " Solute concentration and transport are therefore in]portant considerations in solidification of solutions. The influence of these factors on the dendritic solutions produced from metallic solutions is of practical significance since the properties of the solid aggre-gate are often strongly influenced by dendrite spacing.''' Alexander and Fthines3 found that dendrite spac- ings generally increase with solute concentration in several metallic systems. Michael and Bever1 have observed an increase in a aluminum dendrite spacings in dilute A1-Cu alloys. Rohatgi and AdamsJ have reported a linear increase in a aluminum dendrite spacings with increasing copper contents up to 28 pct in A1-Cu alloys. Bolling and Tiller"' have theoretically examined the effects of solute concentration and diffusivity on the growth of cells and dendrites. Their analysis of dendritic growth assumes a steady-state growth and the maximum possible velocity for dendrite propagation. In addition, the quantitative analysis is restricted to dendrites of rotational symmetry and is therefore not directly applicable to the plate-shaped dendrites of ice obtained in the present study. There are practically no experimental observations of the effect of solute diffusivity on dendrite spacings in metallic systems. The reason appears to be the lack of experimental data on solute diffusion coefficients in liquid metallic solutions. The present work is concerned with the effect of solute concentration and diffusivity on dendritic solidification of dilute aqueous solutions: the solutes examined are potassium chloride. sodium chloride, lithium chloride. and hydrogen chloride. There were several reasons for selecting these solutes. The variation of dendrite spacing with freezing rate in these solutiolls is similar to that observed in A1-Cu alloys.8 These aqueous solutions have almost identical phase diagrams at dilute concentrations. Fig. 1. and the diffusion coefficients show very little variation with solute Concentration.9,10 Dendrite spacings were measured as a function of freezing rate. solute concentration. and solute diffusivity. Dendrite spacings were measured in two different types of freezing systems. droplet and unidirectional, respectively. In the droplet freezing system

Jan 1, 1970