Search Documents

By Frederick C. Hull

THE grain size of a metal or alloy is one of the most important factors determining its properties. In steels, for example, grain size affects hardenability, toughness and machinability; in brasses, grain size is related to hardness and deep-drawing quality. In the field of heat-resistant alloys, grain size has a marked effect on ductility, creep and endurance strength. In order to determine these effects, as well as to control the properties of materials, many accurate and rapid measurements of grain size must be made. This paper presents a new and improved method for the estimation of grain size. USUAL METHODS of EVALUATING GRAIN SIZE A frequently used method of estimating grain size is that of comparing the image of a polished and etched specimen on the ground-glass screen of a metallographic microscope with standard grain-size charts. Several types of standards are available for use with different classes of microstructures. Geometric Grids.-The Metals Handbook1 has a series of hexagonal grids ranging in size from ASTM No. I to 8. Idealized Grain Networks.-Idealized grain-boundary networks for ASTM grain sizes I to 8 are included in the American Society for Testing Materials Tentative Standard E19-39T. Standard Micrographs.- The Metals Handbook and ASTM Tentative Standard EI9-39T contain micrographs of hypoeutectoid and hypereutectoid steels ranging in grain size from No. I to 8. The grain size of annealed brass (diam¬eter of the average grain in millimeters) is obtained by comparison with 10 micrographs of varying grain size in ASTM Tentative Standard Es-39T. A handicap to accurate evaluation of grain size by this method is the fact that it is difficult to compare photographs in a dimly lighted room with the image on the ground-glass screen, or, if the general illumination is increased so that the printed standards may be clearly seen, the contrast of the image of the unknown is decreased. It is easier to estimate grain size by comparing a micrograph of the unknown with the grain-size standards. However, photographing every specimen on which the grain size is desired is too costly in time and materials to be practicable. For hardened steels, the Shepherd or Jernkontoret fracture grain-size standards are very accurate and useful, but this method obviously cannot be applied to ductile alloys such as brass, stainless steels and austenitic heat-resistant alloys. Jeffries' method for grain-size measurements consists of counting the number

Jan 1, 1947

By Ernest N. Patty

WHEN an Alaskan prospector makes a new mineral discovery he stakes out his claims and then starts prospecting for a near-by landing field. This may be a convenient lake but more often it is a gravel bar along a river, where, by cutting off some brush and throwing the larger rocks to one side and then tying a flour sack to a pole, he may decoy some intrepid aviator into testing it gingerly with the wheels of his plane. It was only a few years ago that .the fence around the baseball park at Fairbanks, Alaska, was taken down so that planes could get in and out of the ball park without ripping off boards. Today the grandstand is lost among the modern hangars which line one side of a modern and busy airport. Planes come and go all during the day, and in summer, when there is no darkness, the night stillness is frequently broken by the hum of a motor overhead. Fairbanks is the hub of interior Alaska and from that hub airplanes radiate in all directions to mining camps large and small. An air industry of comparable size is based at the town of Anchorage, and in fact there are but few towns of more than 200 population in the Territory that do not have one or more planes based there.

Jan 1, 1937

By R. L., Fullman

During the past year there have been a number of significant investigations that have furnished evidence on the driving forces governing grain growth and on the role played by boundary impurities. Thus Burke has shown that in high purity alpha brass the rate of grain growth in the earliest stages is approximately inversely proportional to the existing grain size. This relation agrees with one derived by Zener assuming that surface energy was the driving force in grain growth. Also in confirmation of the surface energy hypothesis, Burke found that growth virtually stopped when the grain boundaries or even only the grain corners were melted.

Jan 1, 1949

By L. C. Graton

THE Federal taxes on incomes and excess profits are of course heavy. In 1917, the value of the mineral production of the United States was a little in excess of $5,000,000,000. The total of Federal taxes for that year, for the mineral producing companies, as computed by themselves, amounts, according to tentative compilations, to about $207,000,000, and the total for the comprehensive group of mineral and metal indus-tries to about $766,000,000, which figures represent respectively 9.5 per cent. and 35.1 per cent. of the total tax paid by all corporations. Be-cause of greater complexities in applying the revenue laws to mineral properties than to most other industries, opportunity is afforded for wide differences of view. The taxes indicated above, therefore, are not final

Jan 11, 1919

By Lauren A. Wright, A. E. J. Engel

Under the designations "industrial talc" and "soapstone" are included earth materials of widely different chemical and mineral compositions. Talc, the mineral, is a hydrous magnesium silicate, with a suggested theoretical formula H2Mg3(SiO3)4 or H2O.3MgO.4SiO2, which postulates 63.5 pct SiO2, 31.7 pct MgO, and 4.8 pct H20.18 Single grains or leaves of talc, however, seem to have variations in water content that range from 3.5 pct to more than 7 pct, as well as slight differences in the proportions of MgO and Si02.20 Most of the industrial talcs produced in the United States do not approach the composition given above (see Table 1), but are essen- [ ]

Jan 1, 1960

By J. T. Price

The study was undertaken to investigate the effect in the charge to a blast furnace of zinc and alkali contents, which tend to form accretions and scaffolds in the upper sections of the furnace. Commercial iron oxide pellets were reduced in a CO:N2, 40:60 gas mixture in a 5-cm by 5-m counter-current reactor (Rist-type apparatus) designed to simulate blast-furnace conditions. Pellets from a commercial magnetite concentrate were prepared in the laboratory with zinc (1% ZnO) and alkali (1% KOH) additions to the concentrate. Reducibility, crushing strength and swelling tests were carried out and the results compared with test data from pellets prepared without additions. The results indicated normal behaviour for those pellets containing zinc, whereas those containing alkali swelled catastrophically under the same reducing conditions.

Jan 1, 1977

By T. F. Witherbee

IN the sections, Figures 1, 2, and 3, are shown three crucibles that have been applied to substantially the same furnace, all the conditions having been the same except a variation of one foot of bosh, and a slight variation in the angle of the same. The bosh of No. 3 was originally the same as No. 2, the additional foot in diameter being the result of three months' cutting. Beginning with No. 1, which was 5 feet 6 inches diameter, with tuyeres 5 feet 4 inches high, and cinder-tap four feet from bottom, great trouble was found in holding iron a proper length of time, four hours being the outside limit when working well, and two hours or less if working badly. "Break-outs" through the iron

Jan 1, 1880

By T. F. Witherbee

In the sections, Figures 1, 2, and 3, are shown three crucibles that have been applied to substantially the same furnace, all the conditions having been the same except a variation of one foot of bosh, and a slight variation in the angle of the same. The bosh of No. 3 was originally the same as No. 2, the additional foot in diameter being the result of three months' cutting. Beginning with No. 1, which mas 5 feet 6 inches diameter, with tuyeres 5 feet 4 inches high, and cinder-tap four feet from bottom, great trouble was found in holding iron a proper length of time, four hours being the outside limit when working well, and two hours or less if working badly. "Break-outs" through the iron

Jan 1, 1880

By E. T. Dumble

(Cleveland Meeting, October, 1912.) FOR many years there have been occasional reports of the discovery of gold from a belt of the coast country of Texas which is underlain by deposits belonging to the lower Eocene. For the most part they have been treated as newspaper reports and no examination or record of them has been made. Among such, it may be mentioned that in 1896 there was a flurry over reported gold-discoveries -west of Luling and elsewhere in this belt, and in 1903 some prospect-shafts were sunk on the Gonzales branch of the Galveston, Houston & San Antonio railway, 3 miles below Harwood. During 1912, further discoveries have again brought the matter to our attention, and it may be of interest to record the facts as we have found them. Some years ago specimens were sent me, which were said to have been taken from certain beds on Gazley creek, and concentrated by washing in it rocker. In them were several particles of gold sufficiently large to be visible to the eye. Under the microscope they showed to be, with few, exceptions, sharp cornered, and not rounded as is usual with placer-gold. They were not filings, however, but their sharpness seemed rather to be due to a recent freeing from their matrix. One or two grains were rusty, and one copper-hued, but the most of them were tolerably bright. On examining the material of the samples small grains or fragments were found, of a yellowish brown color, which seemed to be of a finely-granular structure-possibly a sandstone. In these fragments I found particles of gold, one of them being large enough to be visible without a lens. Gazley creek empties into the Colorado river just north of Smithville, and for a few miles north of that place its banks show, beneath the deposits of more recent materials, the sands and clays of the Lignitic division of the Eocene, with lignite, shells of iron, etc.

Oct 1, 1912

By J. F. Keefe, C. J. Bier, O&apos

This study was carried out with copper because it represents the simplest form of sintering, in that but one metal is involved and all reactions occur in the solid state. The present work will cover three variables affecting the sintering process; namely, briquetting pressure, sintering temperature and time in hot zone. Although strength and dimensional changes were observed, this is primarily a metal-lographic study of the sintering process in copper. The term "sintering" should be understood to mean the process by which the metal powder particles of the compact become united into one body. Material Used The copper powder used was made from reduced copper oxide, and is typical of a grade used for the mass production of parts formed of metal powders. The screen analysis of the powder was as follows: Per Cent Per Cent BY BY Particle Size "Weight Particle Size "Weight Mesh Microns /80.... 0.00 43/32.... 17.49 80/I00..... 32/20......16.68 100/140.... 0.34 20/10.... 23.14 140/200___ 15.25 10/5..... 4.64 200/230....... 5.00 5/0...... 0.75 230/325.... 16.70 — 325....62. 69 The minus 325-mesh analysis was made by using a Roller Air Analyzer.1,2 The apparent density is here defined as the weight per cubic centimeter of the loose, unpacked powder as measured with the Hall flowmeter. The value for the powder used was 2.57 grams per cubic centimeter.

Jan 1, 1945

By Frank Firmstone

In 1871, two furnaces at the Glendon Iron Works, which had been blown out on account of the "coal strike," were altered from the open-top plan with side flues for collecting the gas, to closed tops with cup and cone. In one furnace the suppression of the side flues permitted a considerable alteration in the profile, but in the other (No. 5) no change in this respect was made; the side flues were simply bricked up and new ones cut through the sides close to the top. This furnace was in blast over four years, and we can, therefore, form a pretty good idea of the effect of the alterations on the furnace process. The principal dimensions are; Height 72'; greatest diameter, 18'; diameter at tuyeres, 8'; diameter of top, 10'; diameter at stock line (point at which the top of the materials is kept), 11'; cubic contents, 11,900 cubic feet. The table below shows the average results for periods of twentysix weeks each. 1st, of the open-topped furnace; 2d, of the furnace with the first charger used, which was an ordinary cup and cone (diameter of cone 5' 2") 3d, with the charger used ever since, a " double " cone 7' 5" diameter outside, having a central opening 3' in diameter. The diagram (Plate 11) shows the daily variations, in the grade of the irons for the same periods.*

By J. S. Levine, M. Prats

A homogeneous and uniform cylindrical reservoir containing oil and gas is fractured vertically on completion and is produced at a constant bottom-hole pressure. The fracture has an infinite flow capacity, is of limited lateral extent and is bounded above and below by the impermeable strata defining the vertical extent of the reservoir. Results show that such a fractured reservoir can be represented by a reservoir of circular symmetry having very nearly the same production history. The well radius of this circular reservoir is about 1/4 the fracture length and is essentially the same as that obtained previously for a single fluid of constant compressibility. At the same value of cumulative oil production, gas-oil ratios of fractured reservoirs producing at constant terzinal pressure are larger than those of reservoirs having no fractures. This leads to more inefficient use of the reservoir energy in fractured wells and results in lower reservoir pressures for the same cumulative oil production. The reduction in operating life due to fracturing a reservoir is not as great as that for a slightly compressible fluid. This diflerence can be accounted for by the lower reservoir pressure in the fractured reservoir and its adverse effect on the average mobility and compressibility of the oil. As anticipated, the reduction in operating life increases czs the reservoir permeability decreases. The type of results presented in this report can be used to determine the economic attractiveness of fracture treatments per se, to setect the initial spacing to be used in developing a field, and to compare the relative merits of fracturing available wells and infill drilling. INTRODUCTION The effect of vertical fractures on a reservoir producing either an incompressible or a compressible liquid has already been discussed in the 1iterature.l,2 Those results indicate that the production history of such a reservoir is essentially the same as that of a circular reservoir having an effective well radius of approximately one-fourth the fracture length. The present work reports on the effect of a vertical fracture on a reservoir producing two compressible fluids —oil and gas—by solution gas drive. Because of the empirical nature of the PVT and relative permeability data used to obtain the performance of such reservoirs, results can only be obtained numerically and with the aid of high-speed computers. Since reservoirs lose their radial symmetry when fractured vertically, pressure and saturation can no longer be given only in terms of distance from the well. Two coordinates (such as x and y) must now be used to describe the pressure and saturation within the reservoir, and, since we are dealing with compressible fluids, time is also a variable. Thus the solution of a vertically fractured reservoir requires finding two unknowns (pressure and saturation) in two space variables (say x and y) and in time (t). Since no means are readily and generally available for solving such problems at the present time, we have used the results of previous work1,2 to approximate the effect of a vertical fracture on a reservoir producing both oil and gas by depletion. The purpose of the present wmk, then, is to investigate the possibility of using available numerical techniques (limited at the moment to one space variable) to study the two-space-variable flow behavior resulting from a vertical fracture. Results obtained in the course of this investigation are also reported and discussed. Input and output data of the numerical methods used are given in practical units: BOPD, feet, psi, cp, and md. Results are discussed fist in terms of specific reservoir and crude properties and geometries. Later, dimensionless parameters are introduced in order to extend results to different values of some of the reservoir and fracture properties. IDEALIZATION AND DESCRIPTION OF THE FRACTURED SYSTEM It is assumed that a horizontal oil-producing layer of constant thickness and of uniform porosity and permeability is bounded above and below by impermeable strata. The reservoir has an impermeable, circular, cylindrical outer boundary of radius r,. The fracture system is represented by a single, plane, vertical fracture of limited radial extent, bounded by the impermeable matrix above and below' the producing layer (reservoir). It is assumed that there is no pressure drop in the fracture due to fluid flow. 1 indicates the general three-dimensional geometry of the fractured reservoir. Gravity effects and the effects of differential depletion resulting from variations in hydrostatic head (pressure) will be neglected. Thus, the flow behavior in the fractured reservoir is described by the

By AIME

The Taconites Are Ready, the editorial appearing on P. 933 of the September issue, has provoked comment from several informed engineers to the effect that the indicated profit margin was considerably exaggerated. The cost figures in the editorial, taken from the Congressional Record, were characterized as inaccurate by some correspondents. The theme of the story was, however, substantiated in the announcement (P. 1023, October) that Republic and Armco had joined forces in a $160 million project for taconite development. Chief among the dissenters was E. W. Davis, , Director of the Mines Experiment Station at the University of Minnesota. Mr. Davis has worked up the following set of cost figures:

Jan 1, 1950

By Albert Sauveur

(Cleveland Meeting, October, 1912.) ALTHOUGH many metallurgists know that some pearlitic special steels can be made troostitic, martensitic, and even austenitic, without quenching, and, therefore, without exposing them to the dangers of the quenching-bath, the practical significance of such possibility does not seem to have received the attention it deserves. Let us recall briefly the mechanism of the transformation alluded to. In Fig. 1 (after Guillet) is shown the constitutional diagram of nickel-steel that; is, of alloys of iron, carbon, and nickel. Any steel the composition of which falls within the area LPO is pearlitic after slow cooling; any steel the composition of which is represented by a point within the area MLO is martensitic, and any steel within the area MNO is austenitic. Starting with pearlitic steel, then by increasing the carbon-content or the nickel-content, or both, the metal becomes martensitic and eventually austenitic; that is, its composition crosses first the boundary-line LO and then the line M0-and it remains martensitic or austenitic after slow (air) cooling.1 We have inn this way imparted to the metal without quenching it the properties of hardened carbon-steel. In other words, the steel is now self-hardening. Nickel cannot be introduced into solid steel, but by the well-known process of case-hardening it * Professor of Metallurgy and Metalllography in Harvard University. + Assistant in Metallurgy and Metallography in Harvard University. 1 The boundaries between the various zones are not, as a matter of fact, as sharp as indicated in the diagram, because pearlite is not transformed abruptly and in its entirety into martensite when a certain critical composition is attained, nor is martensite suddenly and completely converted into austenite. Greater refinement in the construction of the diagram would undoubtedly result in the introduction of troosto-pearlitic, troostitic, and troosto-martensitic zones between the pearlitic and martensitic areas and-of a martenso-austenitic zone between the martensitic and austenitic areas. Fur the purpose of this note, however, the existence of these transition-conditions may be ignored.

Nov 1, 1912

By Ernest F. Burchard

MOST of the iron ore smelted in Alabama blast furnaces is mined within Alabama, although deposits in the neighboring States of Georgia and Tennessee may be drawn upon when occasion requires. Of the four commercial kinds of iron ore, the anhydrous ferric oxide, hematite (Fe2O3) ; the hydrous ferric oxide, mostly the type commonly known as limonite (2 Fe2O33H2O) ; the carbonate, siderite (FeCO3) ; and the magnetic oxide, magnetite (Fe:3O4) ; the first two are found in abundance in Alabama, the third in less than commercial quantities, and the fourth only rarely. Both the hematite and the hydroxide are found in somewhat different forms according to their geologic environment. The Red Mountain hematite, of Silurian age, that crops out in well¬defined beds for 1.50 miles northeast-southwest and 40 miles northwest-southeast is the predominant source of ore. Only a small part of this area carries ore of commercial value; Birmingham and Gadsden are now the only centers of production. The ore is of sedimentary origin, probably altered from vast deposits of greensand mixed with calcareous, fossiliferous. marine debris. Other types of hematite not vet mined commercially are in metamorphosed rocks-red ore in slates in Columbiana Mountain and specular hematite (gray ore) in quartzite in the Talladega Hills.

Jan 1, 1938

By Stewart, J. B.

THE property at which this work was done consists of a large deposit of potash salts occurring in massive beds of rock salt, overlain by 600 ft. of salt-impregnated shales and marls. It is in the Province of Cataluiia, Spain, about 45 miles from the railway station of Mauresa and about a mile south of Suria. In addition to sinking the shaft, the program of construction work laid out by the European owners (not Spaniards) included a single-track narrow-gage railroad from Mauresa to the mine; one large office building and several large shop buildings, all of brick and concrete with steel truss roofs; two extensive benches for the future recovery plant; and a large village of stone dwellings for employees. The pay-roll showed approximately 1200 to 1500 men employed. After two years' operation, ending early in the autumn of 1920, the shaft had been sunk 325 ft. and still had an equal distance to go to reach the potash; all tunnels for the railroad were done and the road-bed $0 per cent completed; only a little more than the foundations were completed for all the large buildings, and the excavations of the benches were about 75 per cent done-while only a fair start had been made upon the employees' village. This was the state of affairs when the European owners decided to offer this property under a one-year option to an American syndicate, providing that they agreed to carry on the entire building program and to sink the shaft as fast as was consistent with the quality of work which they demanded, for they wished the shaft to endure at least 200 years.

Jan 1, 1926

By AIME AIME

OVER-ALL operations of the oil industry, as measured by production of crude oil and consumption of products, are almost exactly of the same magnitude as a year ago. Does this mean that the great oil industry has been undisturbed by the war? Has this industry, so vital to the prosecution of the war and to the maintenance of our way of life, been unaffected in its operations? No indeed - changes within the industry are great. It has been called upon to revamp its trans¬portation methods so as to move about one fourth of the total amount of crude and refined products by railroad tank car instead of by ocean-going tankers. Its fundamental internal economy, based on maximum gasoline yield from a barrel of crude, has been relegated to memories of the past. Rationing of oil products, which has been forced upon us in face of an abundant supply, is not only disagreeable medicine but is in itself completely disruptive to the marketing end of the-business.-

Jan 1, 1942

G. B. WATERHOUSE,* Buffalo, N. -Y. (written discussion?).-One of the most essential features of rail manufacture is the production of rails that will give good service and be free from failures. To this end rails are made to careful specifications and their manufacture is controlled by rigid inspection on the part of both the producing mill and the railroad or its representative, such as the Hunt special inspection referred to in the paper. Captain Hunt refers to the crescent-shape base failures and describes them as ruptures in the rail flanges extending from the outside inward toward the web. This is hardly correct and may be misleading. It has been shown, over and over again, that the base breaks originating in seams invariably start at the center of the rail base; and run longitudi-nally in varying lengths from 3 to 12 in. and sometimes longer. Most of them are from 5 to 8 in. long and then curve outward to the edge of the flange. The split at the center of the base may exist for an indefinite time and not be discovered until the break shows on top of the flange or the piece of flange breaks off. The. stresses that cause the break are explained in two ways: (1) by the eccentric bearing of the rail base, that is taking contact at one side or edge and bringing excessive strain at the center of the base which starts fracture in any seam that may he present at that point; (2) the split may be due to cross strain in the base where this member is thrown into. compression, as happens where a pair of trucks are on opposite sides of the cross tie or rail-base hearing, and the head is put in tension. A careful study of "split base breaks" shows that fully 95 per cent. occur on the tie bearing, which is further evidence that the above de-scribed causes for the initial break are probably correct. These split base breaks are almost invariably progressive, but always start at the center of the rail-base running longitudinally.

Jan 12, 1919

The Joint Session on Continuous Cast- ing, of the Institute of Metals Division and the Iron and Steel Division of the American Institute of Mining and Metallurgical Engineers, convened in the Jade Room of, the Hotel Waldorf-Astoria, New York City, on Feb. 23, 1944, at 9:15 o'clock, Carl E. Swartz presiding. C. E. SWARTZ.-It is gratifying to see so many out at what may seem an early hour for the beginning of A.I.M.E. sessions, but I think this number-some fifty or sixty of you-is an adequate expression of the interest in these processes all over the United States, and possibly in foreign countries. I would like to ask those who feel that this type of program is of value to you to write or speak to some one of the Institute of Metals Division officers and suggest other topics for symposiums. We have a feeling that some of these current practices are of interest and we believe that under proper sponsorship many of the companies will be willing to give publicity to processes such as those that will be discussed today. Also, we would like suggestions and criticisms, so that the next ones can be better. The subject of continuous casting is not new, of course. The two papers on the program this morning will adequately cover the history and some of the earlier processes, some of the theory and some of the practice, some of the reasons why these various methods did not work. I was rather interested in discovering, when thumbing through one of the A.I.M.E. publications the other night, an article on continuous casting. I was not aware that the A.I.M.E. had published any papers on continuous casting, but there may be a number. The one that I happened to find was published in the 1915 volume, on the Mellen rod- casting process, and so we might say this is not the first session in which these subjects have been under discussion. However, there have been many improvements since then in all fields of metallurgical and mechanical engineering; consequently, I think that the papers today will be a little more involved and a little more complete in their technique. The first paper today will be given by one of our gentlemen of the press, who has done one of the outstanding jobs in this country in calling to our attention some of the newer processes that are being considered and being used in the metallurgical indus-

Jan 1, 1945

By A. W. Fahrenwald

Cell depth has been for many years a controversial question in a flotation-machine performance. In the impeller type of machine, we are really talking about impeller submergence—i.e., the depth in the pulp at which the impeller is required to produce aeration. The Flotation Machine a Compressor A flotation machine can be considered as an air compressor or blower in which the aeration mechanism takes in air under atmospheric pressure† (14.7 Ib. per sq. in.) and blows it into the pulp (the pulp being the receiver) against a head of 14.7 plus cell depth in feet times 0.4333 lb. per sq. inch. In commercial machines the impeller is submerged in the pulp to depths ranging from about 2 ft. in the smaller machines to as much as 5 ft. in some of the larger ones. In the 2-ft. cell, the hydrostatic head on the impeller is (2 X 1.25 =) 2.50 ft. (assuming a pulp density of 1.25), equal to (2.50 X 0.4333 =) 1.078 Ib. per sq. in. For the 5-ft. cell, the hydrostatic head against which the impeller must make air is (5 X 1.25 X 0.4333 =) 2.708 Ib. per sq. inch. In the small 2-ft. machine, the volume reduction is (1.078 ÷ 0.147 + 0.01078 =) 6.85 per cent and in the large, 5-ft. machine, the volume reduction is (2.708 0.147 + 0.02708 =) 15.5 per cent. It is obvious that for unit volume of free air intake the larger machine receives a lesser (8.65 per cent less) volume of aeration* than the smaller machine. The bubbles in escaping from the pulp expand and are greater in volume at the surface than at the moment of their formation in the pulp; 7.33 and 18.4 per cent larger, respectively, in the small and large machines. Just how this situation affects flotation is not known. It is interesting to note, however, that the bubbles are smaller in the zone of the coarsest particles and largest in the upper zones of the her, more easily suspended mineral particles. It would seem that the relationship might be reversed advantageously. The deep cell is at a greater disadvantage in this respect than the shallow cell. Nothing, however, presumably can be done about this situation. The shallow cell is a partial corrective. Not only does the compression ratio of aeration increase with impeller submergence but the hydrostatic head against which the impeller must operate also increases. The Flotation Machine as a Pump Since, in impeller machines that rely upon their own operation to produce aeration, pulp Passes through the im-

Jan 1, 1947