Search Documents

By R. Casting, J. Philibert, and C. Crussard

APPARATUS described in this paper uses the properties of X-radiation, emitted by substances under electron bombardment, as a means of rapid chemical point analysis. The method is based upon the principles of emission X-ray spectrography pioneered by Moseley in 1913. A very finely focused beam of electrons—the electron probe—is projected on the surface of a sample, which has a very large area compared with the beam diameter, at the point where it is desired to know the chemical composition. The minute volume of the sample which is thus irradiated emits a complex X-ray spectrum consisting principally of the characteristic radiations of the elements present in the volume. Measurement of the wavelength and intensity of each component of this spectrum thus affords a simple method of determining the chemical identity and concentration of these elements. The volume of sample which is analyzed at each position of the beam is limited by the diffusion of electrons in the sample. For an accelerating potential of the electron beam of about 30 kv, which is a convenient value for producing the characteristic radiation, this diffusion extends over a diameter of about 1 µ normal to the beam, and about 2 µ in the direction of the beam. Thus a beam of 1 µ in diameter leads to a volume of about 2 µ in diameter being irradiated. The limits of the analyzed region are thus considerably smaller than those obtained by using spark or arc emission spectrometry, which must extend to about 50 µ at right angles to the beam, and those obtained in X-ray fluorescence analysis, where the penetration of the specimen is considerably greater. The apparatus itself consists of three main parts: an electron optical system. an X-ray spectrograph and an optical microscope. The electron opticai system consists of a tungsten cathode emitting the electrow and a pair of magnetic lenses which focus the electron beam on the specimen to be analyzed. The specimen is prepared in the usual way, as in optical micrography. The X-radiation, emitted at the focal spot where the electrons strike the sample, is analyzed with a curved crystal, Geiger-Müller counter, vacuum spectrograph. The light microscope, designed to observe the surface of the sample in the region to be analyzed, is provided with a reflecting objective placed on the axis of the second magnetic lens. Its magnification is X450. The quantitative determination of an element A in the analyzed region is made by measuring the intensity I, of a strong characteristic line of the element, as emitted from the sample, and then the intensity I (A) of this same line as emitted from a

Jan 1, 1958

By E. O. Wagner, V. F. Parry, J. B. Goodman

The low-rank coals containing 10 to 50 pct natural bed moisture represent over half of the tonnage reserve of the available solid fuels of the United States, but only about 2 pct of United States coal production is derived from these fuels. Increased utilization of these large reserves will result if the bed moisture can be removed at the mine in high-speed low-cost processes. This paper presents a theoretical and experimental study of the factors affecting high-speed drying of coal dusts. The time required to heat particles of coal is proportional to the square of their diameter, and it is calculated that the probable time in minutes is equal to 6D2. In other words, a 1/16 in. coal particle should be heated to drying temperature in 1.5 to 2.0 sec. Likewise, 1/8 in. particles should be heated in about 6 sec. The net heat required to dry subbituminous coal and lignite varies from 240 to 375 Btu per pound, depending upon the moisture removed, and this heat can be obtained from 42 to 61 cu ft of hot gases at 600°F. The necessary heat also can be obtained from 7.5 to 11 cu ft of gas at 2000°F. A description of the design and operation of a pilot plant for flash drying coal dusts is presented. This plant dries 1/16 in. by 0 coal dust in suspension in 1.5 to 2 sec with 600°F gas at a rate of 1300 lb per hour per square foot of column area and operates continuously at an efficiency of about 85 pct. The design and operation of another pilot plant for drying 1/8 in. by 0 coal dusts Is also described. This unit dries coal in the fluidized state using the sensible heat of gases introduced at 2000°F and processes coal at a rate of about 1500 Ib per hour per square foot of grate area at 85 to 90 pct efficiency. Over 90 pct of the bed moisture of low-rank coal dusts can be removed by the methods described. The heating value of lignite can be increased 45 pet and that if subbituminous coal 25 pct in a few seconds. The physical and chemical properties of coal dusts before and after drying are presented in the report, and it is shown that the bulk density of the dried coal is equal to that of the raw coal. This phenomenon makes it possible to ship 25 to 45 pct more heat in a railroad car with consequent savings in freight costs. Introduction For several years the Subbituminous Coal and Lignite Section of the Coal Branch of the Bureau of Mines at Golden, Colo., has been concerned with upgrading low-rank coals by removal of natural bed moisture. These fuels contain 10 to 50 pct total moisture, which reduces their industrial value considerably. In 1938, investigations were initiated in cooperation with the University of North Dakota to study drying of lignite by the Fleissner process with heat derived from high-pressure saturated steam. Two pilot plants were built, and two reports were published giving the results of this work.l,2 It was demonstrated that lump lignite and subbituminous coal

Jan 1, 1950

By M. A. El-Aouar

Drilling muds are complex colloidal systems. In an effort to analyze the physical properties that affect their performance at high temperature, an investigation was made of the hydrothermal stability of the monoionic hydrogen, calcium and magnesium forms of montmorillonite and attapulgite. Samples were heated to 400°1F in a simple autoclave and analyzed for cation exchange capacity, c-axis X-ray diffraction spacing and pH factor us sodium hydroxide titration dependence. Out of the six monoionic clay minerals tested, calcium montmorillonite and magnesium attapulgite showed the least base exchange capacities and were relatively unaffected to heat. It is concluded that, of the six clays studied, these two clays were the most suitable forms for use in drilling fluids. INTRODUCTION The drilling of deep wells is dependent upon the drilling mud, and problems frequently arise concerning the effect of mud temperature and prolonged heating. Drilling muds are complex colloidal systems. In an effort to analyze their behavior in deep, hot boreholes, it was thought desirable in this investigation to study suspension of pure, monoionic montmorillonite and attapulgite clays before and after hydrothermal heating. Chemically, clays are hydrated aluminum silicates, such as montmorillonites and attapulgites. The structure of montmorillonite is composed of units made up of two silica tetrahedral sheets with a central alumina octahedral sheet. Attapulgite consists of double silica chains running parallel to the c-axis with the chains linked together at their longitudinal edges. The apexes of the tetrahedrons in successive chains point in opposite direc-tions.4,7,10 One of the most important characteristics of clay minerals is their property of sorbing certain anions and cations and retaining them in exchangeable state. The property of exchange capacity is measured in terms of milliequivalents (meq) per 100 gm of clay. The cation-exchange capacities are taken at pH 7 from titration curves which show the relation between pH and milliequivalents of added NaoH.9,11 The following are causes of cation exchange: (1) broken bonds around the edges of the silica alumina units would give rise to unsatisfied charges, which would be balanced by adsorbed cations; (2) substitution within the lattice structure of trivalent aluminum for quadrivalent silicon in the tetrahedral sheet and of ions of lower valence; and (3) the hydrogen of exposed hydroxyls may be replaced by a cation which would be exchangeable. The primary techniques employed to study the effects of hydrothermal treatment were titration curves and X-ray diffraction analyses. Titration curves were useful in characterizing the nature of the clay mineral surface and the stability of the various samples to attack by acid or alkali before and after hydrothermal treatment. X-ray diffraction analyses were used to determine changes in the nature and crystal structure of the minerals present and to determine the cations in exchange positions on the montmorillonite samples. EXPERMENTAL PROCEDURE PREPARATION OF SAMPLES Selected samples of Wyoming bentonite and attapulgite clay from mines along the Georgia and Florida border were purified by centrifugation to remove all particles larger than 2 microns. In preparing magnesium attapulgite, first a 1.00-normal solution of magnesium acetate was prepared. The pH of the solution was adjusted to 7.0 by adding magnesium oxide. One hundred grams of attapulgite clay were mixed in a blender for 30 minutes with the magnesium acetate solution. The attapulgite solution was left overnight, and then filtered in a Buchner funnel. The suspension was leached with further quantity of magnesium acetate solution and then washed with distilled water for 20 days, after which the filtrate was free from magnesium ions. The concentration of magnesium attapulgite in the clay slurry was adjusted and determined quantita-

By T. W. Lippert

In these opening remarks, I will endeavor to give a background of continuous casting of metals. Only passing attention will be given to the different processes to be described by the authors whose papers are on the program, as they are far more familiar than I am with their subject matter. Suffice to say that over the past several years the brass, copper, aluminum and magnesium industries have shifted practically completely from conventional static molds to continuous casting machines. The Aluminum Company of America alone has more than 60 continuous machines in operation. The product as continuously cast is superior metallurgically, less scrap loss is involved, costs are generally lower, and the equipment requires little floor space. A continuous machine, operating as it does with deadly constancy, is a ravenous consumer of molten metal, and in most plants one or two machines are well able to handle large banks of melting and refining furnaces. In steel, interest in continuous casting is today more serious than ever before in history, and there is every reason to believe that continuous machines may secure a minor place in the ferrous industry, particularly for tool and alloy analyses. In the past, steel has been commercially cast continuously, and today many tons of steel are being turned out in this manner. It has been said that the cradle of every science is surrounded by dead theorists— and certainly there are innumerable examples of this in the metals industries. An excellent illustration, for instance, is the continuous rolling of metal, a tantalizing objective that for decades on end exhausted . the efforts of a long line of famous, infamous and unknown inventors. Finally, in 1922, John B. Tytus, of the American Rolling Mill Co., brought forth what now seems such an obviously simple scheme— a concave roll face, each stand of progressivcly less concavity. And out of this one change in design sprang today's wide continuous mill rolling steel sheet Ioo in. wide at 2000 ft. per minute. Another excellent example is the con- . tinuous casting of liquid metal into an intermediate form, preferably one requiring little subsequent breakdown. First, the casting of liquid metal alone was fascinating enough. Then someone passed the cast metal directly into a rolling mill, and shortly thereafter this continuously rolled strip was passed directly into a stamping machine. This leitmotiv of a continuous flow of metal from the liquid phase to a consumer end product, fast and inexpensive, and all housed within a small space, was enough to drive some 200 or so inventors into frenzy over the past hundred years. All the ideas, however, like mayflies, had only their brief afternoon of existence. They were born and they died on paper, as constant exemplification of that great void that always exists between mental production and commercial production.

Jan 1, 1945

By J. F. Myers, R. J. Tower

"There is nothing new under the sun." All over the world, mineral-dressing engineers are working at their problems, no two of which are alike. Each encounters equipment and process problems. Many devise simple ways and means of overcoming their difficulties. Many of these problems are quite common in the industry, yet, since no one human being seems capable of thinking of everything, these problems at some plants often are looked upon as necessary evils. The purpose of this symposium is to gather together, from competent engineers, devices and practices that have been useful to them. Most of the ideas contributed are not new, yet there is no plant operator, old timer or youngster, who will not find some idea herein that will give him a suggestion for improving his operation. Space does not permit a complete elaboration on all of the contributions offered. Each item is summarized, but this will be ample for those skilled in the art of mineral dressing. A list of contributors appears at the end of the paper, and the superior number at the end of each item identifies its source. General Equipment A signal light to show that a steel bin, containing relatively line, moist ore, is full, can be made by suspending a copper rod in the bin to a sufficient depth to ensure ample contact with the ore. One wire of a 32-volt circuit should be attached to the rod and one to the side of the bin. Sufficient current will flow to make the signal light burn. A 110-volt circuit could be used, but 32 volts ensures safety.10 Fig I shows a circular bin for crushed ore. It is a concrete structure footed on the ground, using no supporting columns. Tailing from the sink-float was used for aggregate. Cross conveyors for discharging ore are not required, nor is a tripper for filling. Segregation is minimized by drawing from different gates. An exhaust fan on top of the bin eliminates condensation due to temperature of the ore.21 A beam with an adjustable point of suspension near the center permits rigging rolls for removal in a very few minutes. This makes unnecessary the lacing of the cable through the hub bolts on one side, around the shaft on the other, and the trouble of balancing the load.8 The ratchet handle on a Symons crusher may be replaced by a drum with a cable wound on it. When the cable is pulled by a crane or hoist it causes the drum to revolve and thus screws or unscrews the bowl. This reduces handwork and materially shortens the time required for the job.' TO Spread ore feed from an apron feeder over a belt conveyor, place a heavy steel bar about midway between the top

Jan 1, 1949

By David White

While a country is small and its coal fields are not many, it may be possible to classify its coals on some basis that avoids both overlap and inconsistency, and that may be satisfactory to the particular country, such as geographic distribution or geologic age, or regional uses. In the adjoining country, however, the coals may differ more or less; broad areas may have escaped the erasure of transition zones by erosion, and the stages of alteration of the coals may lead to a classification different from and even incompatible with that of the other country. Such classifications, developed primarily along national lines, are essentially opportunistic. Further, many even of the "use" terms that have been in more common employment with a degree of international application have lost their significance and value, due to changes in industrial methods and urban life. When the number of coal fields is large and the classification is designed to cover the coals of many countries, the groups or divisions into which the coals are integrated are found to verge one into another; and, though the main divisions may be so naturally composed and oriented that they are relatively distinct and readily separated as entities or units when taken by themselves, the sequence, through classes, groups, and subdivisions, is so complete that we find transition zones through which boundaries must be arbitrarily drawn with minute technical definition. This situation, which is conspicuously illustrated in the United States, is due to the progressive evolution of all coals, irrespective of geologic age, from peats to successively higher ranks, which, if the progressive alteration goes far enough, ends with graphite. This evolution is now firmly established and its complete acceptance is but a matter of time. A classification of all coals in all coal fields can not escape this fact and must take into account all the stages in the transformation of the fuel from rank to rank through division to division, and through group to group. The plan here outlined aims at a natural classification of coals based on their evolution from peats to graphite. As in other classifications, the differentiation of the various groups and divisions takes account of changes resulting from the progressive losses of moisture and .of oxygen of constitution; the relative conservation of carbon; the progressive increase in fixed carbon (inverse to the loss of volatile matter) in

Jan 1, 1930

By Lyall Zickrick

Numerous brands of virgin pig lead are used in the manufacture of lead and lead-alloy cable sheath for commercial cables. In changing from one brand of pig lead to another a readjustment of press operating conditions is nearly always necessary. This is due largely to different flow characteristics, and some brands extrude at a much lower pressure than others; for example, copper-bearing lead requires much more pressure than a low-bismuth lead under the same operating conditions. With the modern type of lead press the extrusion practice in general differs very little. Molten lead at a temperature of 375" to 400" C. (710" to 750" F.) is cast in a water-cooled cylinder and allowed to cool or solidify into a slug for six to nine minutes, This slug is forced into the die block and around the core tube through which the cable runs. The • die block is held at an optimum temperature by means of gas flames, usually 175" to 225" C. (350" to 435" F.). Measurements of the temperature of the cable sheath taken at the time of extrusion have covered a range of 190" to 225" C. (375" to 435" F.). According to Tammann, the rate of extrusion for a metal under constant pressure through a simple orifice is almost doubled for each 10" C. rise in temperature. Because of the irregularity of die-block construction, such a great increase is not obtained in lead-cable extrusion practice. To obtain more definite information about the effect of small percentages of metals usually present as impurities in virgin pig lead, a series of compression tests were made covering several different brands of lead and various alloys. Since the usual temperature of the sheath is approximately 200" C. as it emerges from the die block, these compression tests were all performed at 200" C. Five different brands of virgin pig lead were selected, as follows: Brand A: A desilverized western lead of very high purity. This lead contains neit,her copper nor bismuth. Brand B: A bismuth-free southeastern Missouri lead, usually containing 0.05 to 0.08 per cent copper. Brand C: A copper-free mid-western desilverized lead of very good grade, usually containing around 0.06 to 0.10 per cent bismuth.

Jan 1, 1932

By M. A. Selim, W. H. Somerton

Thermal expansions and heats of reaction of three typical sandstones were measured in the temperature range of 25°to 1,000°C. The significance of these data in subsurface heat-transfer calculations is discussed. Linear thermal expansions were measured both parallel and perpendicular to the bedding planes. Volume expansions are reported as the perpendicular expansion plus two times the parallel expansion. Expansion behavior of the sandstones was found to be controlled by the expansion characteristics of the quartz content. Differential expansion of the quartz grains and other minerals included in the sandstones caused permanent deformation of the heated samples after they were cooled to room temperature. Structural damage resulting from heating is probably an important cause of the reduction of thermal conductivity of heated samples. Measurements of heats of reaction were based on the known heat required for a- quartz inversion. Thermal reactions, which probably include deby-droxylation of clay minerals and decomposition of carbonate minerals contained within the samples, were found to require more than one-fourth of the amount of heat necessary to raise the temperature of the rock alone. In shales and limestones, the reaction heat could be substantially greater than that required from specific heat considerations alone. INTRODUCTION In earlier work, l-3 methods of measuring thermal properties were developed and thermal data for several sedimentary-rock types were reported. Data included thermal conductivity, specific heat and thermal diffusivity. In addition, calculations of heat transfer in subsurface formations may require data on the thermal-expansion characteristics of rock and the heats of reaction of mineral constituents. Thermal expansion of rocks is relatively small in magnitucle and, from the standpoint of change in bulk density, has only minor effects on heat-transfer cliaracteristics. Thermal-expansion behavior, however, may have significant effects on the structure of rocks. Differential expansion of different minerals and along different crystal-lographic axes may result in structural damage which could effectively reduce the thermal conductivity of the rock. Many mineral constituents of sedimentary rocks undergo phase or related changes when heated to sufficiently high temperatures. Although some of these reactions (such as the a-ß quartz inversion) are reversible and the heat absorbed is returned to the system upon cooling, many of the major reactions are irreversible. Through certain temperature ranges, the additional heat energy required to complete the reactions may be nearly as great as that necessary to raise the temperature of the rock in the absence of thermal reactions. If the heat requirements for these reactions are not considered, serious errors in heat-transfer calculations may result. Thermal-tzxpansion and thermal-reaction data have been obtained for three typical sandstones. Methods of measurement and results obtained are reported and discussed in the following. THERMAL EXPANSION OF SANDSTONES The linear expansions of three outcrop sandstones (Bandera, Berea and Boise) have been measured in the temptbratue range of 25° to 1,000° C. Expansion measurements were made in the directions parallel and perpendicular to the bedding. The differential expansion apparatus used is that described by Mitoff and Pask.4 The test sample is heated in an electric furnace with a temperature rise of 6°C/minute. The lengthening of the samples upon heating is compared with the small and known expansion of a fused silica rod. The change in length is transmitted to an X-Y recorder by means of a Statham transducer. A maximum error of ± 1.5 per cent has been obtained with this apparatus

BURKE BAKER, Philadelphia, Pa. (written discussion*).-The small briquetting plant of the American Briquet Co., at 25th Street and Washington Ave., Philadelphia, was built primarily as a demonstration plant. Its output per hour is 4 tons of 2-oz. briquets, which have been retailed in the, vicinity during the year the plant has been in operation. They have proved very acceptable for use in ranges, heaters and grates in-home, and in steam heating plants in apartment houses and hotels. They have not been sold for power purposes, although satisfactory tests have been made in. this field. Anthracite silt is used and is shipped direct from breaker or washery, the only preparation it receives being the removal f excess moisture by means of a Ruggles-Coles rotary dryer. Care is taken, however, to purchase silt containing less than 18 per cent. ash. The method is covered by U. S. Patent No. 941454, issued to Chas. E. Hite, and is known as the "emulsion process." The binder, which is the, basis of the patent, is composed. of 0.5 per cent. cornstarch or wheat flour; 1 per cent. asphaltum or hydrolene, and 6.5 per cent. water, on the weight of coal dust. It is prepared in a tank equipped with paddle agitators and heated by steam jets. The starch and water are first made into a paste by being brought to the boiling point with live steam; the melted asphaltum is then introduced and thoroughly beaten into the paste by the agitators. This rapid stirring breaks. the hot asphaltum into minute particles which are distributed through the paste, producing a smooth chemical emulsion. This emulsion, when dried, is not soluble in water. The binder is mixed with the dry coal dust in a horizontal paddle mixer, 8 per cent. binder to the weight of coal being used. The resultant plastic mass is then discharged into the press, which is of the simple roll type. Since the binder is a liquid and is not sticky or gummy until

Jan 3, 1918

By Charles C. Rodd

PRIOR to 1938, proration procedure in Kansas required the physical testing of wells in order to set up a basis for allocating production. Subsequently the use of liquid-level data and bottom-hole pressure data was authorized on certain types of wells as a basis for calculating well capacities, thereby permitting wells to be tested at low rates and with small volumes of production. This paper gives some attention to the equipment used for obtaining liquid-level data but in general it is devoted to the theoretical and physical factors involved in well testing by drawdown methods. These tests are divided into two classifications: (1) liquid-level data used directly to calculate capacities, and (2) liquid-level data and liquid-level measuring equipment used in connection with other data to determine bottom-hole pressure. Equations and calculations for both types of tests are shown. Well capacities calculated from fill-up data have not been used generally but offer a simple means of testing some wells. Theoretical factors and calculations for such tests are shown. IN the state of Kansas, proration laws required that the "ability of a well to produce" be given consideration in state allocation orders and for many years this was the only factor used in prorating the allowed oil. The "ability of a well to produce" was the quantity of oil that could be obtained from a well in a prescribed period, usually 24 hr. Early in the history of State-regulated proration, Kansas operators realized that, unless some restriction were placed upon equipment, the competition between producers to obtain large well "potentials" might result in a race to install the largest pumping equipment. Consequently, restrictions were placed upon the size of tubing and length of pump stroke that could be used in pumping wells during potential tests. Later, restrictions were placed also upon the size of tubing and chokes that could be used in flowing walls. Despite the restriction set-up, better pumping equipment was designed and installed, pumps were improved, and various devices were invoked to obtain high rates of production for short periods of time. By 1937 pumping cycles of forty to forty-five 54-in. strokes per minute were not uncommon, and numerous wells had produced at the rate of 3500 bbl. per day for short periods. Such production was far beyond any rate that could be long sustained. Equipment necessary to take "potentials" was several times larger than that needed for normal producing purposes, and repairs and replacements required by the high pumping speeds were excessive. Also, many operators were beginning to suspect that the high rates of production maintained during "potential" tests were conducive to early water encroachment. By 1938 large-scale physical testing of wells in Kansas ended, the use of the bottom-hole pressure gauge had become common, engineers and operators were becoming familiar with the term "productivity index," and bottom-hole pressure was being used as a factor in allocation in many areas. There were many drawbacks, however, to large-scale use of

Jan 1, 1942

By Eugene E. Ayres

The centrifuge has come into general use in the oil fields and in the establishments where lubricating and fuel oils are used. Probably centrifugal machinery will be found more useful in the refinery when the operators are more familiar with'the principles of modern centrifuga-tion and when centrifugal engineers better understand the refinery emulsion problems. It is not generally realized that centrifugal force and gravity have two distinct functions: First, subsidence wherein the suspended globules are brought into contact; second, coalescence (when this is possible) by rupturing the films around the globules. In the case of a water-in-oil emulsion, an exercise of the former function alone would yield: (1) oil, free, or as free as desired, from water; and (2) a concentrated water-in-oil emulsion—concentrated in the sense that it contains a relatively large percentage of the internal phase. An exercise of both functions will yield: (1) water-free oil and (2) oil-free water. Centrifugal force is much better than gravity for subsidence because, in the case of industrial emulsions where the suspended globules are visible under a microscope, subsidence is directly proportional to the force applied. Centrifugal force is only slightly better than gravity for coalescence. A minute trace of a proper reagent will immediately do more to nullify the interfacial surface tension opposing coalescence than an indefinite application of the highest attainable centrifugal force. No mathematical analysis has been made to explain the failure of high centrifugal force to induce a decidedly more rapid and more complete coalescence than is obtained by gravity. The adjustment of the centrifuge rotor depends more on the extent of coalescence that is desired or is possible than on any other factor. One reason for this dependence is to be found in the curious relation between the specific gravity and the hydrostatic pressure of emulsions. The continuous centrifuge works on the U-tube or Florentine flask prin-

Jan 1, 1924

By B. G. Matson, M. A. Whitfield, G. R. Dysart

This paper describes the development of u computer program to calculate changes that occur in the length of tubular goods due to temperature and pressure changes during stimulation operations. Due to the numerous variables involved and the uncertainty of all static and dynamic conditions that could exist, it becomes a staggering task for individuals charged with completions to perform the necessary mathematical calculations. The computer program permits advance calculations for several sets of conditions. INTRODUCTION In the Delaware basin of West Texas alone, 50 wells were contracted or drilled to 15,000 ft or deeper in 1965. Deep well activity is continuing in this and other areas on an expanding scale. Many of these deep wells require extensive stimulation for successful commercial production, and during these operations, pressures and temperatures are encountered that have a pronounced effect on the length of tubular goods. This length change during a large-volume, high-pressure stimulation treatment utilizing fluids considerably cooler than bottom-hole temperature can be of such a magnitude that permanent damage to casing and tubing will result unless mechanical design, pressures and fluid temperatures are evaluated and controlled. These pressure and temperature effects can be calculated. However, the process lends itself well to computer solutions because of the mathematical nature of the problem and the calculating hours involved in arriving at an answer. The engineering-hour demand becomes more severe as tapered strings are involved. On initial treatments on a given well, surface pressure and injection rate conditions are unknown, and offset well conditions have not proven to be a reliable method for making predictions. For these reasons, it has become rather standard procedure for operators to compensate for these uncertainties by placing unnecessary pressure and fluid temperature restrictions on stimulation design. On a number of occasions treating fluids have been preheated to as much as 160F as a means of compensating for thermal contmction resulting from pumping cool fluids. The maintenance of packer seals has been treated by Lubinski, Althouse and Logan',' and the problem of therma1 effects on pipe has been explored by Ramey." These works were expanded and the results made applicable to everyday oilfield terminology before submitting them to computer programming. The pressure and temperature effects on tubing movement previously mentioned occur simultaneously as fluid moves through the pipe. The pressure changes, for purposes of explanation, are categorized here as to the various effects these pressures have on a tubing string. These divisions are (1) the piston-like results of forces acting on horizontal surfaces exposed to pressure, (2) swelling or ballooning of the tubing along its entire length due to the forces of pressure acting against the tubing walls, (3) the elongation of tubing due to frictional drag and (4) corkscrewing of the pipe due to internal pressure. Thermal changes are also of great importance, as their results may be more significant than any of the pressure effects. Steel is an excellent conductor of heat and the earth is a relatively poor conductor. It has been calculated that pipe temperatures at depths of more than 20,000 ft approach within as little as 25" the temperfature of the surface fluid after pumping for 2 hours, or a drop in temperature in some treatments of more than 220F. The equations presented in this paper were developed for computer programming and simplicity of input information; therefore, numerical constants such as Young's modulus for steel (28 X 10\ si), the coefficient of thermal expansion of steel (6.9 X 10."IF) and Poisson's ratio for steel (0.3) are included with unit conversion factors. The moment of inertia of tubing cross-sectional area with respect to its diameter was changed to a constant times (D' — d') where D is outer diameter and d is inner diameter. Units in the equations are length in feet, diameter in inches, density in pounds per gallon, pressure in psi, rate in barrels per minute and time in hours. PISTON-LIKE REACTIONS A change in tubing internal dimensions and the exposure of other horizontal surfaces to different pressures on the inside and outside of the tubing result in a reaction much like a piston under pressure. Such is the case when the internal diameter changes in a combination string of pipe, when seals of a slick joint assembly are subject to pressure and in the end effects of a tubing string. The change in tubing length due to the piston effects of a slick joint packer is affected by the various diameters involved, the tubing pressure Ap,, the casing pressure ,Ap,, length of pipe L, densities of fluid in the tubing before and during pump-

By T. Egleston

circumstances, would prefer the steel with which they are now familiar, to a specimen that Mr. Sandberg has described as having broken into seventeen pieces under the wheels. After blowing such low manganese steel, it may be coaxed into a rail, but it is a wonder that it holds together so long as it does, with so great a nornber of minute flaws. I would not in any way depreciate chemistry, but I think it should be kept in its proper sphere. Let the chemist look after the quality of pig metal, and apply common sense in the avoiding of extremes, then the most fastidious railroad cannot find fault with the result. PROFESSOR EGLESTON, New YORK City : It is not my place as an engineer to apologize for the chemists, but as no one seems disposed to do so, and as they have had more than their share of criticism, 1 am glad to say that I believe there are chemists in this Institute whose work and word are just as reliable, arid perhaps even more so, than that of the average engineer. But we ought to make a distinction; there are chemists and chemists. With the ordinary commercial chemist, who looks upon the science as so much merohandise, I have not a particle of sympathy; but with the chemist who looks upon his profession as engineers do upon theirs, I have every sympathy. When matiufactorers and engineers go to a chemist and ask him to make an investigation, and scretv him dowu to the lowest point, turning the equivalent of his brains into cents and mills, they ustially get an exact equivalent in poor work for miserable pay, and no one has or should have any sympathy with them, and the under these conditions has no right to ask for any, and no reason to criticise any work that he may get under such circumstances. But I am disposed to think that the chemists who have been replwented and discussed at this meeting do their work conscientiously, and that it is as reliable as that of almost any profession. I believe, however, that this problem of steel rails is being investigated in a wrong direction. I said so at the Pittsburgh meeting, and I think the discussion of this meeting will prove it to all those who have heard it. I think the chemigt is incapable of solving this problem unlees he goes very far into the domain of physical chemistry, so far that it becomes physics and not chemistry; and that the physicist will be the one on whom we must rely in the future for the elucidation of the subject. The chemist may aid the physicist, but it is my decided opinion that we are looking in the wrong place to get the explanation of the phenomena.

Jan 1, 1881

By V. S. Tuman

In the first part of this paper, an estimate is made of the magnitude and extent of the thermal stresses which result from mud circulation. Our study is made for the period of relaxation, i.e., when the drilling operations are terminated and mud circulation is discontinued. In the second part, we have shown how the thermal stress can affect the wave propagation. An example is worked out for a Carbondale sandstone (Illinois basin). It is concluded that, in extreme cases when differential temperature is about 100°F, the travel time will be affected by about 2 microsec/ft. For more refined permeability studies, such corrections would be desirable. MAGNITUDE AND EXTENT OF THERMAL STRESSES RESULTING FROM MUD CIRCULATION When geological beds are penetrated by boreholes, it is observed that the formation temperatures increase with depth; however, the temperature profile varies considerably from place to place. During drilling operations, mud pumped from a pit at atmospheric temperature absorbs heat from the lower beds. As a result, the drilling fluid returning to the surface is at a higher temperature than the surrounding beds. The temperature gradient in a well depends on the rate of mud circulation and other factors; normally, this temperature gradient is smaller than the natural gradient in the formations. (See Fig. 1.) The temperature difference between the mud column and the surrounding beds gives rise to therma1 stresses (which are also important in the phenomenon of fracturing), and it is expected that they will affect the transit time of the sonic front in velocity logging.1 It is evident chat beds near the surface will undergo various cycles and will have a complicated dynamical and thermal-stress history. A limestone bed which is just penetrated will act as a source of heat, warming up the cold mud pumped down a few minutes earlier from the surface pit. The same limestone bed a week later will act as a sink, absorbing heat from the drilling mud which has already warmed up sufficiently to be at a higher temperature than the bed in question. No doubt a large section of the hole will undergo a heat treatment, which will make the calculation of the exact magnitude of the thermal stresses very difficult. In the solution of wellbore thermal equation, it is assumed that the rock around the wellbore is elastic and isotropic, and that the system is radially symmetrical. A complete dynamical solution of the thermal stresses set up at different horizons during the drilling operations would require a comprehensive history of the mud circulation, the surface temperature of the pit and the number of hours in which mud was not circulated. CONDUCTION OF HEAT AND THERMAL EQUATION OF STATE

By Fred C. Bond

COMPRESSION TESTS THE standard method of determining the crushing resistance of rocks consists of crushing prepared shapes under slow compression, and expressing the ultimate crushing resistance at the load causing failure in pounds per square inch of crossectional area perpendicular to the crushing force, with the height approximately equal to the diameter. One-inch cubes or cylindrical drill cores are commonly used. Cubes and drill cores are cut by a high-speed steel disk with diamond dust embedded in the edges, and running under a water spray. Such a saw will cut one square inch or more of hard stone per minute. An oiled, spherical, swivel compression block of small diameter should be used to equalize the pressure. Each sample is crushed between sheets of cardboard or blotting paper extending about one-half inch beyond the edges of the stone. A cloth is wrapped around it to prevent dangerous explosive shattering. It is probably preferable to place the swivel block below the specimen, with either a rigid head or another swivel block above. Whatever the arrangement, the conditions should be selected that will develop the maximum crushing strength of the specimen, even though its top and bottom planes may not be precisely parallel. The cardboard pads may be dispensed with if the ends of the specimens are polished smooth. The use of lead pads was discontinued because of the observed tendency to split the material. The samples are weighed and measured before breaking, and the density is calculated. Cubes that contain bedding planes or veinlets are usually placed with these vertical and parallel to the crushing force, in order to develop the maximum resistance possible. A number of compression tests were made on Canadian ores, by Forrest Nagler, of the Allis-Chalmers Manufacturing Co., and are listed alphabetically in Table 1. These tests were made on one inch cubes or on diamond-drill cores approximately one inch in diameter and one inch long. The drill cores represent virgin rock, while the cubes may have been previously weakened by explosives. Compression tests have been made in the Allis-Chalmers Milwaukee Laboratory on a number of ores and other materials. These are listed alphabetically in Table 2. In test No. 1339, on limestone from the Hanna Coal Co., comparison was made between 2 by 2-in. diamond-drill core segments and one-inch cubes cut from the cores in the same strata. Results for the cylinders are tabulated directly below those for the cubes. The bedding planes were normal to the compressive force in breaking the cylinders and parallel in breaking the cubes, which accounts for the lower strength obtained from the cylinders. The percentage of net linear compression at the ultimate pressure was measured in test No. 1055 for Fontana TVA quartzite as 0.30 for cube A, and 0.37 for cube D.

Jan 1, 1946

By C. G. Goetzel

Until recently porous bronzes have found many applications for self-lubricating bearings in the automotive, electrical, household appliance and general machine industries. The bulk of an annual production of several million pounds of copper powder had gone into the manufacture of these articles, of which by now more than a billion are in service. However, the war has brought severe curtailment of these bearings because of the scarcity of the two main constituents, copper and tin. The development of substitute materials of equal or superior properties has led to a large-scale replacement by products made from iron powder. In spite of the diminished interest in powdered copper-tin alloys caused by this trend, it was believed worth while to investigate some mechanical properties of these materials as a reference base for substitutes. In addition, experiments using the hot-press method were incorporated to investigate possibilities of saving vital equipment or opening new fields of applications. It must be assumed that by this time the powder metallurgy process of making porous bronzes, a development that found its first industrial applications more than 10 years ago, is well known in engineering. The patent literature and bibliography is volu- minous, most publications referring to porous, self-lubricating bronzes, containing about 10 per cent tin and I to 5 per cent graphite, with porosities in the range of 25 to 35 per cent.l-7 The excellent antifriction properties of these bearings, caused by the uniform impregnation with oil, are dependent on the interconnection of the pores and the uniform distribution of the graphite inclusions. The sintering mechanism of porous bronze, with its structural changes during the transformation of the metal powder into the sintered bearing has been shown in colored photomicrographs by Hall.* Diffusion between copper and tin was found to begin within one minute, when tin-rich transitory phases were formed. After 25 min. a homogeneous matrix of alpha was reached for an alloy containing 9.5 per cent tin and 6 per cent graphite. Price, Williams and Garrand,9 studying certain metal powder combinations in which sintering is accompanied by the cementing action of a liquid phase, including a 90-10 bronze sintered at 925°C., have found that if the treatment is carried out within the liquidus and solidus temperature range of the system, a peculiar process of alternate solution and precipitation occurs. The particles of the major solid constituent are dissolved in the minor liquid phase, but thereafter are reprecipi-tated on certain nuclei, which develop into large grains, usually rounded.

Jan 1, 1945

Discussion of the paper of W. R. Shimer and F. O. Kichline, presented at the New York Meeting, October, 1913, and printed in Bulletin No. 81, September, 1913, pp. 2361 to 2377. ALLERTON S. CUSHMAN,* Washington, D. C.:-I have been much interested in Mr. Shimer's valuable contribution. The apparent fact that the oxides of certain metals which are not reduced by hydrogen per se, are reducible when the oxides are dissolved in iron in small quantity, is brought out for the first time, and is worthy of the careful attention of investigators. The Ledebur method has been subjected to considerable criticism, owing to the alleged fact that it did not measure the oxygen combined with impurities. Personally I have found the method extremely helpful in spite of the minute quantities of oxygen with which it deals. Mr. Shinier points out that the highest oxygen he obtained by over-blowing was 0.074 per cent., while the minimum quantity in a properly finished steel was in the neighborhood of 0.02 per cent. The Ledebur method is used as a constant check on the quality of heats made in practice under my direction. In dealing with a very pure commercial iron, we should expect to find and would find difficulty in galvanizing sheets made from a heat running 0.07 per cent. oxygen. In fact, we do not like to have the oxygen rise above 0.03 per cent. I am inclined to agree with the author that excessive oxygen leaves a bath of metal in a very short time, but nevertheless I believe it takes only a very little combined oxygen to begin to make its presence felt in the pickling and galvanizing department, when sheet metal is the product manufactured. As Mr. Shimer points out, the higher the carbon, the lower the oxygen danger, and it is partly for this reason that the manufacture of carbon-free metal in the open-hearth furnace has to be so very carefully watched and controlled from the furnace through to the finished product. It is quite possible that, there is a critical temperature of dissociation between iron and oxygen, although the critical points may vary in baths of different carbon content. Unfortunately, the accurate determination of the temperature of a molten bath at the time of pouring or teeming is a very difficult matter. For a long time I have been wishing that there were some reliable data which would indicate the effect of the tem-

Jan 12, 1913

Several of the current concepts of brittle fracture involve consideration of, the existence of defects or flaws. Griffith 1 observed that the tensile strength of brittle materials was several orders of magnitude less than might be expected from the calculated strength of the molecular bonds. He was able to show that the stress concentration at the tip of an elliptical crack is sufficient to raise the local stress level many fold. He therefore proposed that real materials do contain such stress risers in the form of minute cracks and that this accounts for the rather large difference between theoretical and actual strengths of brittle materials. The degree of stress magnification depends to a great extent upon the orientation of the crack with respect to the general stress system, and to the general geometry of the crack. It is but a small step to extend the theory to include flaws and defects of other sorts which, though they may be less tractable mathematically than cracks, may still function as stress risers in a real sense in real materials. Brace 2 proposed that the application of the Griffith theory to rock material could be tested in two ways: "(a) by finding the orientation of cracks from which macroscopic fractures have developed and (b) comparing failure strengths from tests at different conditions of loading." In 1963 the author and coworkers at the Bureau of Mines, Twin Cities Mining Research Center, working on a problem involving relationships between strength, sonic velocity, and microstructure "eveloped an experimental technique to examine several somewhat similar questions; namely. does the frequency of occurrence of actual or potential stress risers dictate the relative strength of the material as proposed by Weibull;4 can actual flaws or defects in the rock structure be identified and related to strength and perhaps other properties in accordance with Griffith's theory; and is the anisotropic strength and elastic behavior of certain rocks related to the preferred orientation of these actual or potential stress risers? The experimental approach developed to answer these questions appears equally applicable to the testing of Griffith's theory as proposed by Brace. Basically the approach consists of a series of point-load tensile tests followed by petrofabric examination of thin sections prepared from the specimens in order to identify and apply a quantitative measure to microstructural defects which may have contributed to the failure process. In the point-load test, tensile stresses are induced by a compressive load applied to the center of a disc-shaped specimen by means of two opposing hemispherical indentors. Failure generally occurs as a single fracture parallel to the axis of loading and bisecting the disc. Because of the geometrical symmetrical the induced tensile stresses are symmetrical about the loading axis

Jan 1, 1967

By R. I. Jaffee, F. C. Holden, H. R. Ogden

ECENT papers dealing with the properties of unalloyed iodide titanium have been directed primarily at the determination of base-line properties for alloy investigations. Early work was limited to a few tests because of the limited availability of iodide titanium at the time. In the results of papers by Campbell et al.,1 Gonser and Litton,2 Jaffee and Campbell,3 inlay and Snyder,4 and Jaffee, Ogden, and Maykuth, data on mechanical properties are presented for unalloyed iodide titanium in the annealed and cold-worked conditions. Data are presented in this paper which show the effects of heat treatment on the structure and mechanical properties of commercially produced iodide titanium. Correlation is made between microstruc-tural variables and the mechanical properties. Experimental Procedures Melting Stock: The melting stock used was as-deposited iodide titanium, produced by New Jersey Zinc Co. The furnished analysis showed the following range of impurities: N, 0.004 to 0.008 pct; Mn, 0.005 to 0.013; Fe, 0.0035 to 0.025; Al, 0.013 to 0.015; Mo, 0.0015; Pb, 0.0045 to 0.0065; Cu, 0.0015 to 0.002; Sn, 0.001 to 0.01; Mg, 0.0015 to 0.002; and Ni, 0.003. Hydrogen content as determined by vacuum-fusion analysis was 0.0091 wt pct (0.44 atomic pct) after arc melting and fabrication. Nitrogen analysis on the arc-melted and fabricated titanium showed a content of less than 0.002 pct N. The average hardness of the furnished stock was Rf 70, or approximately 85 VHN. Melting Procedure: The as-deposited rods were rolled, sheared, and degreased in preparation for arc melting. The charge was arc-melted with a tungsten electrode in a water-cooled copper crucible under a positive pressure of high purity (99.96 pct) argon. The final ingot was approximately 2 in. in diameter and showed no increase in hardness over that of the initial stock. Fabrication: Heating for fabrication was done in air. It was begun by forging the ingot into a 3/4 in. diam rod, at an initial temperature of 1600°F. Scale was removed by sandblasting. The rod was then swaged to 1/4 in. diam at room temperature through a series of 20 dies, with approximately 10 pct reduction in area between each die. An anneal of 1 hr at 850 °C in air was given after the 1/2 in. die, such that the final cold reduction was 75 pct. Sections cut from this rod were used for test and microstructure specimens. Heat Treatment: Heat treatments were carried out in resistance tube furnaces with stainless-steel linings, under an atmosphere of gettered argon. As further protection against contamination, the specimens were packed in titanium turnings in a titanium sleeve. Control experiments have shown negligible hardness increases with this method, indicating that contamination from oxygen and nitrogen is slight. Three cooling rates were employed in this work; these have been designated as water quenching, argon cooling (to simulate air cooling under a controlled atmosphere), and furnace cooling. The cooling rate for an argon cool is 100°C per min for the first minute, with an average cooling rate of 35°C per min over a 15-min period. A furnace cool requires about 10 hr, with an average cooling rate of 3.6oC per min during the first hour, and an average cooling rate of 1.2°C per min over the 10-hr period. Microimpact Test: The specimen adopted was based on the cylindrical Izod Type Y specimen (ASTM, E23-41T). All dimensions were reduced to half scale, including the notch radius. Specifications are shown in Fig. 1. The specimen is held vertically in an adapter and broken as a cantilever beam. Impact tests were run on a constant-velocity (11.34 ft per sec) Tinius Olsen impact testing machine with a total available energy of 100 in.-lb. Tests were made to determine the correlation between this microimpact and the standard V-notch Charpy impact test. Curves showing impact energy as a function of temperature for both impact tests are plotted in Fig. 2. Transition temperatures, when they occur, are about the same for both impact tests. All three titanium-rich materials have the same conversion factor, 10. Tensile Testing: Tensile tests were conducted on Baldwin-Southwark testing machines using the 600, 2400, or 3000 1b range. Specifications for the test specimen were taken from the 1948 edition of the ASM Metals Handbook, and are shown in Fig. 1. Strain measurements were made using an SR-4 resistance gage (Type A-7) cemented to the reduced section in conjunction with a lever-type extenso-meter. Readings on the SR-4 strain indicator were

Jan 1, 1954

By R. N. Hoffman

A new technique for determining capillary pressure curves has been developed and tested. The technique differs from previously reported centrifuge techniques in that the centrifuge is slowly accelerated from zero to the maximum desired speed rather than being held constant at particular, progressively higher speeds. An important advantage of this technique over other methods of determining capillary pressure curves is the short time required to obtain the desired amount of data over the chosen pressure range. For example, to obtain 30 data points between 1.2 and 104 psig with a 1.55-in. long, 3/4-in. diameter core using a brine-air system, 6.6 hours were required with this technique. An equally important development of this paper is an analytic method for the conversion of the data from the centrifuge experiment to capillary pressure curve data. Previously there has been only an approximate conversion available. Although the capillary pressure curves determined by this technique appear to be as accurate as those determined by other techniques, the accuracy could be improved if certain variables, not treated in this experiment, were investigated. Among these are the dynamic distortion of the centrifuge equipment and imperfect initial saturation of the cores. INTRODUCTION Pirson 1 defines capillary pressure in a porous medium as "the differential pressure that exists between two fluid phases at their interfaces when they are distributed under static equilibrium within a porous material". Capillary pressure in rocks is known to be a function of fluid saturation, among other things, and a capillary pressure curve is defined for the purposes of this paper as a plot of the capillary pressure-wetting phase saturation relationship for a particular rock sample. Several methods are used for determining capillary pressure curves for small rock cores. Prominent among these are the semi-permeable barrier, mercury injection, and, to a lesser extent, centrifuge methods. The semi-permeable barrier method is currently the most popular. It features simplicity in both execution and the mathematical conversion of the experimental data into a capillary pressure curve. The main disadvantages of the semi-permeable barrier method are the time required — as long as two months — to obtain several points for the curve, and a fairly low maximum pressure before breakthrough of the non-wetting phase into the barrier occurs, for example, about 32 psig for a brine-air system. It is for these reasons that other methods such as the centrifuge method have been introduced. High accelerations and the absence of a barrier result in quicker attainment of saturation equilibrium at a given pressure. However, the centrifuge method involves much more expensive equipment and more difficult procedures and calculations than does the barrier method. The purpose of this investigation has been to improve the equipment and procedures of the centrifuge method and to develop an analytic method for the conversion of the experimental data into a capillary pressure curve. Hassler and Brunner did the original work in the determination of capillary pressure using a centrifuge. In their work the centrifuge speed was increased in a step-wise manner, each speed being held constant until saturation equilibrium was reached in each core. Saturation equilibrium was indicated when the volume of liquid collected in the graduated pipette of the core holder remained constant. According to Hassler and Brunner, equilibrium was reached in "a few minutes to one-half hour or more". In the centrifuge method, as opposed to the barrier method, the fluid saturation of the core is not a constant throughout the length of the core, but varies with the radius of centrifugation. Also, the capillary pressure cannot be read directly but must be calculated from a knowledge of the centrifuge speed and other parameters. For these calculations, Hassler and Brunner derived the following equations: