Search Documents

By R. C. Edwards, T. G. Kirkland, Will Mitchell

A scrubber of new design has been invented for the beneficiation of glass and chemical sands. The machine is described and its capacity and metallurgical efficiency compared with the performance of other sand cleaning devices. THREE years ago, when the Process Research Laboratory at Allis-Chalmers Manufacturing Co. sought a remedy for the increasing cost of disposing of great quantities of spent sands from foundries, R. C. Edwards developed a continuous inclined paddle scrubber which removed a substantial amount of the charred coatings from the grains with a surprisingly small amount of degradation of the sand itself.' The satisfactory results obtained on foundry sand, using this piece of equipment in conjunction with classification, suggested that a similar procedure could probably be applied to the cleaning of other sands, such as those required by the glass and chemical industries. This paper describes the flowsheet and the results of tests performed on impure natural sands. The four materials tested were silica sands obtained from widely separated geographical areas. Three of the sands tested contained various amounts of iron oxide and the fourth an excessive amount of calcareous bonding material. The iron oxide was present predominantly in the form of grain coatings, which were scuffed by the paddle scrubber, and in all cases a product was manufactured which would meet one of the sets of specifications imposed by the glass industry.' The Edwards paddle scrubber used in these tests consists of a rubber-lined steel cylinder 18 in. in diam and 48 in. long, mounted on a frame so that its slope can be varied 5" to 35" from horizontal. This variable slope permits controlled flow of material through the machine. Rubber paddles rotate inside the cylinder and are secured to a center shaft 2 in. square, which extends through the entire length and is turned at both ends so that it can be supported by sleeve bearings centrally fixed in the end plates of the shell. At the upper or feed end it extends through the bearing on the end plate into a sheave, to which the motor drive is connected by means of Texrope transmission. The four sets of paddles mounted on the shaft, two in each set, are made of rubber belting, 1/4 in. thick, 9 in. wide, and 16 in. long. The two paddles in each set are secured on opposite sides of the square shaft, but in the same longitudinal position along the shaft axis. In other words, the two paddles in the set operate 180" apart. Each successive set is mounted 90" from the preceding one. One end of each paddle is bolted to the square shaft and the other weighted by a steel bar, lx1 1hx9 in. long, which is attached about Yz in. from the outer tip of the paddle. By centrifugal force at proper shaft speed, these weights force the paddle surface to rub smoothly against the rubber lining of the cylinder. During operation, the sand forces its way between the paddle face and the surface of the cylinder, providing the necessary abrading action as the pulp flows by gravity down the length of the cylinder. The capacity of the machine is limited by the amount and the hardness of the coating material and by the quality of product desired. The degree of scrubbing is governed by the retention time in the scrubber, which in turn may be changed by varying the feed rate, water content of the feed, and

Jan 1, 1953

By V. Carlson, J. A. Rowe, E. O. Bennett

This report concerns the microbial flora found throughout the surface facilities of six water-injection systems in Texas and Oklahoma. Each system is described in detail and water quality data are presented when available. Pseudomonads and sulfate-reducers (Desulfovibrio) were the predominate organisms found in the systems. The genera Sphaerotilus, Bacillus, Achromobacter, Minococcus, Clostridium, Flavobacteriurn, and Sarcina were also isolated in significant numbers. Molds, iron bacteria, sulfur bacteria and soil bacteria were found less frequently in the systems. It is concluded that a complex microbiological flora exists in oilfield water-injection systems. A tendency for waters to deteriorate in passage through solid bed filters is also noted. INTRODUCTION The injection of water into a subsurface formation is often accompanied by numerous problems, the origin and nature of which are only partially understood. During the past few years, investigators have indicated that bacteria are involved in some of these problems even though very little specific information is available concerning this matter. It is surprising that investigators working on microbiological problems in water injection have never made a complete study of the flora of these systems. The logical course to pursue is that of identifying the microbial flora and then examining each organism individually and in combination with others to determine if they produce a particular problem. The objective of the present investigation is to identify the microbial flora in water-injection systems. In addition. the quality of the water in most of the systems was determined and some observations made on the extent to which passage through surface-handling facilities affects water quality. The information derived from this study can serve as a foundation for additional basic investigations. EXPERIMENTAL PROCEDURE The microbial flora of six water-injection systems were studied at 50 sampling points. These systems are experiencing problems which may be attributable to micro-organisms, and two were being treated with antimicrobial agents at the time of this study. The samples were obtained by inoculating 1.0 ml of the water onto the different media at the sampling point. The media used for primary isolation included nutrient agar, Strokes agar, M-10-E medium and semi-solid thioglycollate medium. The cultures were returned to the laboratory, incubated at 30°C until growth occured; then, pure cultures were obtained from individual colonies. The pure cultures were identified according to standardbacteriologica1 procedures. Those organisms not fitting identification criteria exactly, even though exhibiting close similarities to known species, are listed as unidentified organisms. Membrane filters were used to make water-quality tests at intervals before and after collection of the microbiological samples. The filter discs were 47 mm in diameter with a 0.45-micron particle-size retention, and the tests were done under closed conditions at 20) psi. Membrane flow rates are plotted as increasing filtrate volumes, and the total solids contents arcs given when available. An effort was made to sample at each place in a system where a change in environment occurred. In this manner, the effect of passage through filters, tanks, hay sections, etc., on the microbial flora and water quality could be observed. RESULTS SYSTEM A This waterflood system (Fig. 1) is about four years old and daily pumps approx&ately 110,000 bbl of oil-free sour water from a 3,700-ft deep reef. The

By James F. Watson, Jack L. Christian

The austenite to martensite solid state phase transformation in the 300 series austenitic stainless steels was studied over the temperature range of 788° to -423°F. The primary variables controlling this transformation are temperature, plastic strain, and chernical composition. Consequently, temperature effects were studied by measurements at 78", -100°, -320°, ad -423°F; plastic strain was controlled by cold-rolling the alloys varying amounts between 40 and 80 pet; and chemical compositions were varied by using 301, 302, 304 ELC, and 310 stainless steel alloys which exhibit varying degrees of austenite stability. The martensite contents of various alloys were correlated with several mechanical properties including yield and ultimate tensile strength, elongation, and notched tensile strength (Kt = 6.3). The yield and tensite strengths generally increased with increasing martensite content, but ) no correlation was observed between elongation and martensite content. The notched/un-notched ratio (Kt = 6.3) at -423°F exhibited a maximun at about 65 pct martensite. Attempts to correlate austenite stability (as measured by the Ms temperature) with both martensite content, and incremental increases in martensite content resulting from tensile testing were generally unsuccessful because the variable of cold Work could not be separated. As a result, only a general correlation of these variables was obtained. A (111) pole figure of austenite in a cold-rolled 301 stainless steel mas prepared an attempt to further investigate directionality as exhibited by mechan- ical properties. Yield strength was correlated with this pole figure, but ultimate and notched tensile strengths were not. 1 HE literature contains a wide variety of references pertaining to the austenite to martensite solid state phase transformation in austenitic stainless steel, see references 1 through 18. Discussions of the crystallography and kinetics of this transformation are given in references 2, 7, 8, 'and 11, while further discussions of austenite stability and hardening mechanisms are given in references 1, 3, and 10. Two important processing variables involved in this reaction are the degree of cold work and the temperature at which cold work is performed, and the nature of these effects are outlined in references 4 and 12, and 15 and 16 respectively. The latter two references deal primarily with cold rolling conducted at subzero temperatures. The effect of chemical composition on the austenite to martensite reaction was treated in detail by Eichelman and Hull, and they determined an equation for the M, temperature of this type of steel which included the effects of nickel, chromium, carbon, nitrogen, silicon, and manganese.5 Previous studies have shown that austenitic stainless steels of AISI types 301, 302, 304ELC, 310, and 301-N exhibit mechanical properties between 78" and -423"~ that are microstructure dependent, and the dependence of this microstructure on chemistry, temperature, tensile stress, and prior deformation is in general accord with previous investigations which have measured the effects of each of these variables on structure, or more specifically, the austenite-to-martensite transformation.lg The behavior of fully stable alloys, such as ATST type 310, at cryogenic temperatures is

Jan 1, 1962

By J. M. Smith, D. Kunii

Effective thermal conductivities of sandstones filled with stagnant fluids were measured using a steady-state technique. Data were obtained for seven sandstone samples, taken from four different locations and ranging in permeability from 18 to 590 md. The measurements with gases (helium, nitrogen, air and carbon dioxide) covered a pressure range from 0.039 psia to 400 psig. Data were taken for four liquids — n-heptane, methyl alcohol, 79.8 weight per cent glycerol-water solution and pure water at atmospheric pressure. The experimental results were used to evaluate the theoretical equations for predicting stagnant conductivities developed earlier 4 The low-pressure measurements permitted evaluation of the consolidation parameter bpDp/ks (necessary to utilize the theory) for the various types of sandstones. Using these characteristic values, the tbeoretical equations correlated well with the experimental conductivity data for the several fluids and rock samples. INTRODUCTION An aspect of heat transfer in solid-fluid systems of considerable current interest is the effective thermal conductivity of porous media. The stimulus for study of the subject arises from the need for sound procedures for designing thermal methods of petroleum production. The general system occurs when there exists a flow of fluid through the pores of the solid material. However, a logical starting point in developing a theory for predicting the effective thermal conductivity in the general system is to attack the special case when the porous solid is filled with stagnant fluid. Since the flow rates anticipated in thermal production processes are very low, such stagnant conductivities k,O are also of practical significance. Kunii and Smith4 recently reviewed the existing data for stagnant conductivities and proposed a theory for heat transfer in porous rocks. This leads to equations for predicting the stagnant conductivity as a function of the properties of the fluid and solid phases. The existing experimental information1-3,5,7,8 covered a range of porosities and agreed well with the theory. However, the available data were insufficient to examine critically the effects of two important properties on koe, — the thermal conductivity of the fluid and the pressure. The objectives of the present study are twofold: (1) to measure the effect on koe of (a) the pressure for gases, and (L) the thermal conductivity of the fluid kg for bloth gases and liquids; and (2) to determine for rocks of different porosities and permeabilities the apparent solid conductivity ks and the consolidation parameter bpDp/ks required in applying the theoletical equations. Measurements extrapolated to zero pressure are helpful in evaluating ks and the consolidation parameter. However, the extrapolations must be based upon data obtained at low pressures because in small pores the molecular conductivity of the fluid may be unusually depressed. This occurs when the mean free path of the molecules is of the same order of magnitude as the pore diameter, that is, in small-diameter pores at low pressures. The phenomenon has been studied recently by Schotte6in beds of unconsolidated fine particles. With particles in the size range of 45 to 200 microns, a significant reduction of kg with pressure was found even at absolute pressures greater than 1 atm. In the sandstones employed in the present investigation, pore diameters lower than those experienced by Schotte are likely, so that the effect would presumably be greater. SCOPE OF MEASUREMENTS Stagnant conductivities were measured for four types of sandstones using a total of seven samples. The characteristics of the samples, including porosity and permeability values, are given in Table

By W. H. Triplett

DISCUSSION John G. Barry (Mexico)—"The Nerinea limestone," writes W. H. Triplett, p. 584, "is ... medium-bedded (0.5 to 1.0 m)...." This is the Zuloaga formation. It has always been described as thick-bedded (1.0 to 3.0 m). "Whether the marls and shales were less amenable to replacement," he goes on to say, ". . . is a matter of opinion." In general, it is probable that the thinner-bedded formations were less competent than the nearby thicker-bedded, so that the latter produced pre-mineral breccias when all were subjected to flexing stresses. With reference to Mr. Triplett's analysis, p. 584, under Lower Cretaceous, in the Taraises and Cupido formations some few horizons should be considered as possibly containing hidden orebodies. In the Potrero camp, shown on the general map in the northwestern part of Fig. 4, p. 587, several mineralized beds are known; they are found throughout the whole thickness of the Cuesta del Cura formation. With regard to intrusive rocks, p. 585, the general map, Fig. 4, shows that the Potrero and Noche Buena camps are near a third, monzonitic intrusive. It seems doubtful that the Aranzazu was any hotter than the Providencia camp. However, the former does contain large deposits of exogenetic silicates, which in large part occur within the Zuloaga limestone, as shown by Bergeats in a stereogram. At the Aranzazti camp the La Caja formation was only slightly changed, whereas it was altered highly at the Providencia, possibly en-dogenetically. That the temperatures at the Providencia camp were relatively high at one time is also shown by the tourmaline veinlets and lenses in the monzonite and by the garnets on Temeroso peak; high temperature mineralization has possibly been eroded. Maps and sections show that the Concepcion del Oro —Temeroso stock occupies the original locus of some 40 kms of sedimentary rocks. The remains of large roof pendants at Carmen and Animas indicate that stoping was active. Whether all the sedimentaries were stoped and assimilated presents a lively problem, a correlative of which would be an estimate of the original constitution of the invading instrusive rock. As to extrusive rocks, see p. 585, there is also a large thick area of the extrusives left near Noche Buena in the "La Caja" intramontane valley of the La Caja range. This place gave its name to the range. The volcanics are also said to occur in considerable amount in the Mazapil valley.'' Turning now to the remarks under Regional Structure, p. 586, as partly indicated on the general map, Fig. 4, the Mazapil Mountains of the Mazapil region are a continuous, U-shaped range including a northern leg of some 24 km, a southern leg of similar length, and a uniting base which lies to the East and which measures some 10 km in a S 20 E direction. Thus the U-shaped mountains are continuous through a length of 58 km and also contain one continuous anticline, the adjacent valleys being synclinal. The northern leg of the U is called the La Caja range, the base is called the Concepcion del Oro range, and the southern leg the Santa Rosa range. The ranges in northeastern Zacatecas and southern Coahuila contain anticlines and the valleys contain synclines, which where thin are necessarily highly compressed isoclines with very steep dips. Many of the anticlines also have isoclinal components which are overturned, original mushroom tops being indicated. The anticline in the base of the U of the Mazapil Region, Concepcion del Oro range, is a recumbent anticline, over to the ENE, with a necessarily subjacent recumbent syncline.

Jan 1, 1953

By G. W. Nabor, H. J. Ramey

A blotter-type electrolytic model was utilized to prepare flow diagrams for a field test of the in-situ combustion process. It is pointed out that the areal sweep of a combustion pattern is similar to sweep patterns that would be developed at an infinite mobility ratio, which exists (approximately) across a combustion front because of the complete removal of liquids from the sand behind it. The precision of the blotter method was tested by comparison with results obtained by other techniques and was found to be satisfactory. The blotter-type model will not furnish as much information as more elaborate and expensive potentiometric models, but its speed of operation and ease of construction make it a highly satisfactory tool to determine areal sweep patterns. A tabulation of sweep efficiency and mobility ratio is furnished for various well geometries. INTRODUCTION In connection with the first field experiment with the in-situ combustion method of oil recovery," flow diagrams for the unique well geometry and boundary conditions involved in the subject test were developed. From these diagrams, information concerning the areal sweep and distribution of flow was obtained. A survev of the literature on potentiometric models, utilized to solve two-dimensional flow problems,1,2,3,4,5,6,8,9,10 indicated that the information desired could be obtained by means of a blotter-type model by making certain modifications in the technique. The following discussion describes the application of the modified blotter-type electrolytic model to the solution of the problem of the flow of gases in an oil reservoir for an engineering appraisal of the progress of the combustion front during the first field experiment. Considerable information has been published on the theoretical aspects of electrical model techniques. One author stated that:' "The models are based on the observation that since the velocity of an ion in an electrolytic system is proportional to the potential gradient, just as the velocity of a fluid particle in a porous medium is proportional to the pressure gradient, the paths of the ions in the electrolytic systems must be similar to those of the fluid particles in a porous medium with the same geometry and equivalent boundary conditions." This suggestion of a real analogy has been established repeatedly by comparisons of analytic derivations with model results. The analogy between the movement of a combustion front and the model results is not as apparent, since the combustion process differs from water flooding and gas cycling in that the velocity of the burning front is different from the velocity of the injected fluid at the front in combustion. Analogy between electric current flow and movement of the burning front exists because the local velocity of the burning front is proportional to the air (mass) velocity at this point (assuming frontal temperatures are at least as high as the ignition temperature of the fuel). Most model studies fall into one of two categories: (1) electron conduction, or (2) ionic conduction models. Examples of these are models which employ as flow media either a thin metal sheet, or a porous blotter saturated with an electrolyte, respectively. Of the two, the porous blotter model, an ionic conduction model, appears to be the most adaptable to the flow problem that results when using in-situ combustion. This problem requires that the mobility of the particles in the model be controlled analogous to the mobility of gas flowing through burned sand of high permeability to gas and then through oil sand of lower permeability to gas. Although the blotter model has heretofore been

Jan 1, 1955

By H. B. Heller, T. J. LaChapelle

Phosphorus nitride has been used as a diffusant for introducing phosphorus into silicon under various conditions. It has a temperature -dependent rate of decomposition beginning in the 500°C range, increasing very rapidly above 1000 C, producing gas -eous phosphorus and nitrogen. Best control was obtained with a two-temperature arrangement, with silicon at the proper higher diffusion temperature. It is an easier compound to use than the usual P2O5 because it is nonhygroscopic and yields temperature-controlled surface concentrations together with excellent final surfaces. Surface doping nearly as high as that produced by phosphorus pentoxide has been observed for sources at 850" to 900°C. Reproducible high sheet resistances were obtained by using slow flow rates (1 linear in. per min) and staggered-baffle gas mixing with the nitride at 550°C. Both inert gas (argon, nitrogen) and oxidizing ambients have been examined. Oxygen reacts to produce anhydrous P2O5, consuming the source at a rapid rate. Nitrogen, except for short diffusion times, suppressed the dissociation reaction and sometimes produced inferior surfaces. Argon appears to be a more useful ambient. PHOSPHORUS is a common impurity atom which may be introduced into a semiconductor lattice to create a p-n junction or an nn contact region. Various investigators1-4 have described techniques which can be used with either elemental phosphorus or certain of its compounds. This paper discusses a newly available compound of phosphorus, phos- some processes silicon wafers are precoated with a phosphate glass which acts as the doping source. Although a nonoxidizing form like elemental phosphorus or a phosphine can be used, most diffusion methods usually involve the use of phosphorus pentoxide. Because of the rapid volatility of this compound, a two-step process is often used, combining a gaseous deposition to introduce a predetermined quantity of phosphorus into the silicon with a later redistribution to obtain the desired distribution and surface concentration. Such a process usually involves predepositing phosphorus at surface concentrations greater than 1020 cm-'. At such doping levels the phosphorus-diffusion coefficient appears to be strongly concentration-dependent.596 Distribution profiles5'9 for these high concentrations of phosphorus do not follow Fick's laws, having extended subsurface regions which appear electrically to be nearly uniformly doped with ionized impurity atoms. Radioactive-tracer experiments also show near-surface anomalies in the distribution profile. Additionally, it has been shown1' that the introduction of very high surface concentrations of impurity atoms into silicon (a step necessitated by the subsequent lowering of concentration during the redistribution) creates dislocations, even in material having no dislocations originally. For device fabrication the use of oxide masking introduces certain difficulties and limitations.7'8 The standard practice of using phosphorus pentoxide as an impurity source adds other problems to these because of the speed and ease with which this compound reacts with moisture present in any gas in contact with it. In fact, P2O5 is one of the fastest and most effective drying agents for gases. Because it is such an excellent desiccant, completely anhydrous phosphorus pentoxide is difficult to obtain and is even more difficult to keep dry. Adequately dry gloved boxes must be used for storage and transfer of the chemical. The presence of even traces of moisture will cause formation of phosphoric acids which, during the high-temperature diffusion, results in polyphosphoric acids of variable composition. Such reaction by-products end up as gradually accumulating viscous fluids in the cooler parts of the diffusion furnace. The calculation of the final solute distribution for

Jan 1, 1964

By H. C. H. Darley

Particles of montmorillonite-type clays consist of plate-like clay crystals stacked face to face. Under certain conditions in a suspension, these particles disperse to individual platelets with the formation of diffuse double layers round each platelet. This action has a profound influence on the rheo-logical and other properties of the suspension. This paper describes a test for degree of dispersion and uses this test to correlate changes in dispersion with changes in viscosity and gel strength. The test is of practical value in determining and controlling dispersion in drilling muds. INTRODUCTION It has long been recognized that the hydration and dispersion of clays, particularly those of the montmoril-loilite type, strongly influences the rheological and other properties of drilling muds, but little has been done to measure either of these phenomena under conditions that exist in drilling muds. The hydration of clays as such has been extensively studied by those interested in clay mineralogy, but in drilling muds it is to be expected that dispersion is at least as important a factor as hydration because of the much higher water-to-clay ratios and the very high degree of mechanical agitation. The term dispersion as generally applied to colloidal systems means the subdivision of large particles to small. Particles of montmorillonitic clays, however, consist of plate-like particles stacked face to face. Dispersion of these aggregates to individual platelets may be expected to have an unusually large influence on the properties of the suspension because of the shape factor, the enormous increase in specific surface, and the pronounced surface phenomena exhibited by the crystal surfaces. It is this form of dispersion that is particularly considered in this paper, -and the term degree of dispersion is used to imply the number of platelets per particle. The term aggregation is used to denote the opposite of dispersion, the association of individual platelets face to face to form stacks. It is important, as Garrison' has pointed out, to distinguish between the deflocculation-flocculation phenomenon and the dispersion-aggregation phenomenon. Flocculation of clay suspensions is characterized by the formation of voluminous sediments that are stable under quite high centrifugal fields. This indicates an edge to plate association of particles to form an open network type of structure, which obviously will have quite a different influence on the properties of the suspension than aggregation of the platelets face to face. The work described in this paper was primarily concerned with developing a test for degree of dispersion in clay suspensions and by means of this test to follow changes of dispersion brought about by addition of electrolytes and to show the influence of dispersion on the rheological properties. The Clay-Volume Test This test is based on the assumption that the greater the number of particles per unit weight of clay in a suspension the greater will be the volume of sediment when it is flocculated under standard conditions and centrifuged. After considerable experimentation the following procedure was established: 1 cc of the test mud vigorously shaken with 100 cc of 1-N NaCl in an ASTM oil centrifuge tube and centrifuged for five minutes in an International No. 2 centrifuge at 2,000 rpm. When drilling fluids of different solids content are being compared, the clay volume is expressed in cc per gram of clay solids, but generally changes in dispersion in a given mud are being compared and it is more convenient to report as cc of sediment. The test is strictly a comparative test, so that it is not necessary to adhere to the conditions specified above, e.g., with field muds a stronger flocculant such as Al2(SO4)3 may be preferable to NaCl.

Jan 1, 1958

By C. Perez-Rosales

A simplified method for evaluating specific surface of porous media is described. The theory takes as its point of departure a previous statistical method. Due to its simplicity, it is considered that the proposed method has certain advantages over methods currently used. By making some simplifications, the procedure for estimating specific surface is reduced to superposing, on an enlarged reproduction of a section of the sample to be analyzed, an evenly spaced grid and counting the number of intersections of the grid lines with the perimeter of pores. An indirect test of the theory is presented and the advantages of the method are discussed. INTRODUCTION The specific surface of a porous medium, defined as the surface area of the pores per unit bulk volume, is an important parameter with regard to flow of fluids through porous materials. In particular, it has long been recognized that specific surface is closely related to permeability. Since the internal surface of any porous material is extremely complex, the measurement of specific surface can only be accomplished by indirect means. It is considered that the best method' in current use for determining specific surface is the statistical method of Chalkley, Cornfield and Parka who originally applied it for estimating volume-surface ratios of morphologic components, such as cells or nuclei. This method will be outlined later. A method which has been used extensively in determining specific surface is based on the adsorption of a vapor by the internal surface of porous media. However, this method does not yield the surface pertinent to fluid flow, since it includes the tiny molecular interstices which do not influence the permeability of the porous material.' This method has been employed by Brooks and Purcell measure internal surface area of sedimentary rocks. The Kozeny equation, which relates the specific surface to the permeability of a porous medium, has found wide use in determination of specific surface. However, the KOzeny equation is not strictly correct. Large discrepancies have been found between theory and experiments.' Consequently. the specific surface estimated by this method is subject to serious uncertainties. Finally, there are other methods for specific surface measurements which are based upon a variety of physical and chemical properties: the heat conduction by a gas, ionic adsorption phenomena and direct chemical reactions of the surface with other substances. This paper describes a simplified statistical method for determining specific surface of porous materials. It is considered that, due to its simplicity, it has certain advantages over the methods presently used. For its validity, it is required that the porous structure of the material under study be of a spatially random nature. This is not a serious limitation since the naturally occurring porous media have this characteristic. THEORY The method proposed in this paper takes as its point of departure the interesting result obtained by Chalkley et al The method shows that if a line of length r is dropped repeatedly at random over a section of a porous medium, and counts are made of the number of times the two end points fall within pore areas (denoted by h for "hits") and the number of times the line intersects the perimeter of pores (denoted by c for "cuts"), then in a very large number of throws where S is the internal surface of the sample (surface of pores) and V, the volume of the pores (Fig. 1). Since the specific surface S, and the porosity 4 are given, respectively, by S. = Sl V, and 4 = V,l V,, where V, is the total volume of the sample, it follows from Eq. 1

By P. D. Southgate

VIBRATION, both ultrasonic and sonic, can affect the course of a variety of metallurgical processes. Reports of work on this subject have appeared at intervals over the last twenty years and a thorough review has been made by Hiedemann,' who gives a full list of references. It has been established that both refinement of structure and outgassing of the melt can be produced in most metals with sufficiently intense vibration. Many of the experiments in this field have been carried out using alloys of direct commercial interest in production- type casting equipment. Here a different approach is adopted, and simple binary alloys which exhibit different types of structure are used to facilitate examination of the effects produced under well-controlled conditions. Results from this kind of experiment should enable the basis of larger scale experiments with similar types of materials to be better substantiated. The alloys selected for this work are: 1) aluminum of 99.2 pct purity, the principal impurity being iron, near 0.4 pct; 2) A1-1.8 pct Fe, which is near eutectic composition; 3) A1-4 pct Fe—the second phase, of composition FeAl3, forms distinct acicular crystals; and 4) A1-12 pct Si—although this is of eutectic composition, the silicon normally appears in the form of large plates. Experimental Method There are three practicable methods by which acoustic vibration can be transferred to liquid metals: a) vibration of the whole of the container; b) electromagnetic induction; and c) insertion of a vibrating probe into the liquid. Method c is chosen here, and a frequency of 8 kc per sec is used. The end of the probe is electrically heated so that the metal near it will be the last to solidify. This ensures that vibration is transmitted directly to the liquid metal without passing through any solidifying and pasty intermediate region, which would attenuate acoustic waves severely. In addition, the heating prevents solid metal adhering to the probe, with consequent detuning. The transducer assembly is shown in Fig. 1. A stack of Permendur laminations 30 cm long acts as a half-wave vibrator and a winding along its length carries polarizing and energizing currents. It is joined by a half-wave coupling bar to a mild steel exponential horn, which tapers from 3.8 to 1.1 cm diam. A stainless steel stub 3 cm in length is welded on the free end. The assembly is supported by a plate located at the node of the coupling bar. Acoustic radiation losses are minimized by allowing the free end of the transducer to protrude above the surface of the cooling water. The exponential horn serves the function of matching the acoustic output impedance of the transducer to the radiation resistance of the probe into the melt. Electrical drive is provided by a 1 kw generator. Measurements of probe vibration amplitude during melt treatment are obtained from a gramophone pick-up with the needle attached to the wide end of the horn. This device is calibrated with a microscope before use. Liquid aluminum is very actively erosive and solution of materials with which the molten metal comes into contactcauses difficulties. Erosion of the probes in particular can be severe since it is greatly accelerated by the vibration. Tests of various acous-

Jan 1, 1958

By G. Thodos, R. B. Grieves

A method has been developed For predicting the critical temperatures and critical pressures of binary mixtures of carbon dioxide, hydrogen sulfide, nitrogen, hydrogen, carbon monoxide and oxygen with the normal paraffin hydrocarbons. For carbon-dioxide and hydrogen-sulfide systems, relations are presented that take into account the peculiar behavior of mixtures with closely boiling components, such as carbon dioxide-ethane and hydrogen sulfide-propane mixtures which exhibit minimum critical temperature points. For hydrogen, nitrogen and carbon-monoxide systems, the extreme critical behavior caused by wide differences in pure component properties is established In addition, those fixed gas-paraffin systems which resemble paraffin-paraffin systems are also accounted for. For a mixture of known composition, the pure component critical temperatures, critical pressures and normal boiling points are all that are required to determine its critical point. Graphical relations are presented relating Tc and PC , of the mixture to the pure component properties. From the treatment of 12 carbon-dioxide and hydrogen-sulfide systems reported in the literature (74 mixtures), the expected error lor the critical temperature is approximately 1.5 per cent, and for the critical pressure, approximately 2 per cent. From the treatment of six hydrogen, nitrogen and carbon-monoxide systems reported in the literature (30 mixtures), the expected error for both the critical temperature and critical pressure is approximately 2.5 per cent. The relationships, which have been developed with only normal paraffins as the hydrocarbon components, may be extended to those isoparaffins and olefins which fall within the allowable volatility ranges. INTRODUCTION Many of the fixed gases — carbon dioxide, hydrogen sulfide, nitrogen, hydrogen, carbon monoxide and oxygen — occur in natural mixtures with hydrocarbons. Carbon dioxide and hydrogen sulfide are frequent components of the fluids produced from underground petroleum reservoirs. Nitrogen, carbon dioxide and hydrogen sulfide are present in varying quantities in most natural gases and gas-condensate well effluents. Hydrogen mixtures are of considerable interest in many phases of refining processes of petroleum. The determination of the critical temperatures and critical pressures of such mixtures is of value in vapor-liquid equilibrium studies, for the prediction of the characteristics of underground reservoirs, and for reduced-state correlations of PVT, transport and thermodynamic properties. The accurate estimation of the critical point for binary mixtures is an important initial step toward a complete analysis for the establishment of the critical temperatures and pressures of multi-component mixtures. Methods for predicting the critical temperatures and critical pressures of binary hydrocarbon systems have already been presented in the literature.8 It is possible to apply these existing methods to fixed gas-paraffin mixtures, but due to their unusual critical behavior, values calculated deviate considerably from experimental values. For systems containing trace quantities of the fixed gases, these methods are acceptable; however, for systems containing more than 5 mol per cent of the fixed gases, these utterly fail to produce reasonable critical values. Consequently, in this study a method has been developed for handling such binary mixtures over the entire composition range. CARBON-DIOXIDE AND HYDROGEN-SULFIDE SYSTEMS The critical behavior and the vapor pressure behavior of mixtures of carbon dioxide and hydrogen sulfide with paraffinic hydrocarbons may be quite similar or quite dissimilar to that of paraffin-paraffin mixtures, depending on the volatilities of the components involved. The critical temperature and normal boiling point of carbon dioxide are very close to the corresponding values for ethane, while its critical pressure is considerably higher than that of ethane. The normal boiling point of

By M. Hansen, C. C. Wang

THE 6 phase of the Cu-Sn system (Cu,Sn, = 32.53 pct Sn) was considered to be stable down to room temperature until Owen and Iball showed by X-ray analysis that it undergoes a eutectoid decomposition into the a and e phases at a temperature near 300°C. Their findings have been confirmed by other investigators,2-5 a11 of whom used X-ray diffraction studies of powdered specimens. The transformation temperature has been reported by various authors as lying between 300" and 380°C. To our knowledge, no microscopic evidence of the eutectoid reaction has been published, as yet. It was in connection with other work that longtime annealings at 300°C were conducted with alloys containing 22, 32.6, and 35 pct Sn (nominal compositions) prepared from electrolytic copper and electrolytic tin and cast in a small iron mold. The specimens were first annealed at 700°C for 5½ hr and quenched from 400 °C after cooling to, and annealing at, that temperature for 9½ hr. After this treatment the sample with 22 pct Sn was cold worked for 15 pct reduction in thickness; the 32.6 and 35 pct alloys were too brittle to be cold worked. The final anneal at 300°C was done in a sealed pyrex glass bulb for 63 days. Fig. 1 represents the microstructure of the annealed alloy with 22 pct Sn. It shows primary a solid-solution crystals in a matrix of the (a + 8)-eutectoid formed at 520 °C. The a crystals contain 8 inclusions which precipitated from a due to the decrease in the solubility of tin in copper with fall in temperature. The decomposition of the 8 phase into a (light) and E (dark) can clearly be seen in some areas of the (a + 8)-eutectoid matrix. However, the decomposition is far from being completed, even after the long-time anneal used. Figs. 2 and 3 depict the microstructures of the annealed 32.6 pct alloy, which was intended to consist entirely of 6 phase. Due to some loss in copper on melting, the composition is slightly higher in tin. Accordingly, the sample after quenching from 400 "C consisted of grains of the 8 phase with the e phase arranged along the grain boundaries and as a Wid-manstatten precipitate within the 6 grains. After an annealing time of 63 days, the eutectoid decomposition, 6 + a + e, is still in its early, initial state. The pearlitic type of eutectoid is nucleated at the c crystals, preferably those surrounding the grain boundaries of the 8 phase. The appearance of the (a + €)-eutectoid in the 35 pct Sn alloy was very similar to that shown in Figs. 2 and 3. Again, the eutectoid decomposition nucleated at the 6 phase and proceeded from there into the 8 areas. As is evident from Fig. 2, the rate of decomposition of the 6 phase at 300°C is extremely sluggish: In the representative area shown only about 5 pct of the 8 phase has undergone the eutectoid trans- formation after annealing for 63 days. On the other hand, Owen and Williams2 have reported that X-ray photographs of powdered samples with between 17.5 and 34 pct Sn did not show 8 lines after annealing for only five to seven days at 300°C. This is in accordance with the fact that the rate of reaction in powdered samples is higher than in massive specimens, as a result of the excess free energy of the powder (due to its greater surface area). The authors wish to express their thanks to C. A. Johnson for preparation of the micrographs.

Jan 1, 1952

By H. A. Deans

A three-parameter mathematical model for one-dimensional flow in porous media is developed. The objective of the model is to predict accurately the longitudinal dispersion associated with the flow of either gases or liquids over a wide range of Reynolds number. A qualitative analysis of the model parameters is given. Published experimental pulse-response curves are compared with results predicted by the model. Several general types of problems are given for which the model can be used. INTRODUCTION The problem of multicomponent single-phase flow through porous media is encountered in the study of petroleum reservoirs, gas chromatographic and ion-exchange columns, industrial fixed-bed con-tacters and elsewhere. These particular examples span the wide range of flow conditions possible; Reynolds numbers of less than l0-4 are not unusual in oil-production problems, while values in excess of 103 are common in large fixed-bed operations. The study of flow-dependent transport phenomena is complicated both by the changes in the character of the flow over this range, and by the irregularity of the flow boundaries inherent to porous media. Dispersion is one of the important phenomena known to depend fundamentally on flow conditions as well as on fluid and medium properties. As used in this paper, the term "dispersion" refers to the observed mixing of fluid elements of different composition which occurs in flow systems. The actual mechanism may be one or more of a number listed below. Only dispersion in the direction of the mean flow (referred to as axial or longitudinal dispersion, or mixing) is considered here, although lateral dispersion arises as part of certain coupled mechanisms. MECHANISMS FOR LONGITUDINAL DISPERSION A number of distinct mechanisms are known3,5,7,8 to contribute to the phenomenon of longitudinal dispersion. The more important of these are as follows. 1. Molecular diffusion in the flow direction. 2. Turbulent (cell) mixing. 3. Lateral transport processes coupled with velocity and/or residence time distributions, including: (a) "Taylor" diffusion caused by the interaction of velocity profiles in individual voids with lateral molecular diffusion; (b) separation and remixing or interdiffusion of streams having different velocities around particles; and (c) the couplipg of gross velocity profiles, caused by viscous instability or inhomogeneous porosity, with lateral dispersion. 4. Finite mass-transfer rate between a porous matrix and the flowing phase, and finite diffusion rate inside elements of the porous matrix. It is a well documented fact that these mechanisms, whether acting individually or in combinations, produce essentially similar integral effects. The observed dispersion can be described approximately by solutions of the diffusion equation with a properly chosen dispersion coefficient. Necessary conditions for equivalence between dispersion and diffusion are discussed by Klinkenberg and Sjenitzer.7 Apparently these conditions are not satisfied exactly in many real systems. "Best fit" solutions of the diffusion equation often depart appreciably from measured breakthrough curves for the usual pulse- or step-forced experimental systems. In particular, pulse response curves usually show some degree of asymmetry and "tailing" which cannot be reproduced by the one-parameter diffusion model. The three-parameter model proposed herein attempts to produce these effects in a realistic manner. Asymmetry and tailing are predicted for certain sets of parameter values, while for other sets the new model agrees with the diffusion model. VELOCITY DEPENDENCE OF DISPERSION Molecular diffusion apparently controls at sufficiently low Reynolds number in both gas and liquid systems.*l3 At Re > 20, turbulent cell mixing is dominant for gases flowing at low pressure; the data of McHenry and Wilhelm8 show that the mixing cell is approximately 1 particle diameter long in random sphere packs. on the other hand, the data

By W. A. Backofen, B. B. Hundy

THE pole figure interpretation in a recent paper' on the torsion texture of copper was questioned in discussion,2 nd a simpler alternative interpretation was proposed. In the hope of reaching a positive decision in favor of one or the other, further studies were undertaken with 70-30 brass and Armco iron. This additional work has not answered all the questions about torsion textures, but the new pole figures do offer support for the original interpretation and show that the texture is more complicated than the alternative suggests. From the previous paper,' the appropriate method of describing the texture in a twisted bar is known to be a pole figure plotted on a plane tangent to the surface of the bar. The axes of the pole figure are chosen to coincide with directions in the tangent plane which are parallel and at right angles to the bar axis (the longitudinal direction, L.D., and the transverse direction, T.D.), for these directions define symmetry axes of the texture. Therefore any ideal orientation that might be selected to represent the texture gives the crystallographic plane parallel to the tangent plane and the crystallographic direction parallel to the axis of the torsion specimen. A unique texture was assumed in the original interpretation, and the four ideal orientations, (112) [111], (111) [112], (110) [001], and (112) [131], were used in its description. Of the four orientations, the first three provide possible slip directions along the direction labeled T.D. in the pole figure, which is one of maximum shear stress in a specimen being twisted. The alternative, however, simply considered the texture to be made up of two components, one a [Ill] alignment along the specimen axis with a random orientation about the axis as in the wire texture, and the other a superimposed orientation (110) [0011. Experimental evidence for this simpler interpretation was derived from two observations: 1—twisting did not alter the orientation of a single-crystal torsion specimen of a brass in which an octahedral plane was perpendicular (and a [Ill] direction parallel) to the axis, and 2—deformation bands with approximately the (110) [001] orientation were seen to develop in the crystal during the test. The new pole figures were obtained with a brass and an iron torsion specimen of 3/4 and 1 in. gage length diameter, respectively. After annealing, polishing, and twisting, seven rods, about 0.030 in. in diameter and with different orientations, were prepared from material near the surface of each specimen; these were required for the method of constructing pole figures by using a Geiger counter as developed by Norton.5 Since the shear strain gradient in plastic torsion is linear, it was readily computed that the material supplying the brass rods had been subjected to a shear strain of about 3.2, and that supplying the iron rods had received a shear strain of about 2. All details concerning specimen design, testing procedure, rod preparation, and use of the Norton technique can be found in ref. 1. Only this procedure was followed in working out the (111) and (100) pole figures for brass, Figs. 1 and 2. However, X-ray photograms were also involved in the construction of the (110) pole figure for Armco iron, Fig. 3. The photograms were made according to the conventional transmission diffraction technique using a small sheet taken parallel to the specimen surface and etched to a thickness of about 0.0025 in. They were especially important in locating the boundaries

Jan 1, 1954

By R. C. Edwards, Will Mitchell, T. G. Kirkland

A scrubber of new design has been invented for the beneficiation of glass and chemical sands. The machine is described and its capacity and metallurgical efficiency compared with the performance of other sand cleaning devices. THREE years ago, when the Process Research Laboratory at Allis-Chalmers Manufacturing Co. sought a remedy for the increasing cost of disposing of great quantities of spent sands from foundries, R. C. Edwards developed a continuous inclined paddle scrubber which removed a substantial amount of the charred coatings from the grains with a surprisingly small amount of degradation of the sand itself.' The satisfactory results obtained on foundry sand, using this piece of equipment in conjunction with classification, suggested that a similar procedure could probably be applied to the cleaning of other sands, such as those required by the glass and chemical industries. This paper describes the flowsheet and the results of tests performed on impure natural sands. The four materials tested were silica sands obtained from widely separated geographical areas. Three of the sands tested contained various amounts of iron oxide and the fourth an excessive amount of calcareous bonding material. The iron oxide was present predominantly in the form of grain coatings, which were scuffed by the paddle scrubber, and in all cases a product was manufactured which would meet one of the sets of specifications imposed by the glass industry.' The Edwards paddle scrubber used in these tests consists of a rubber-lined steel cylinder 18 in. in diam and 48 in. long, mounted on a frame so that its slope can be varied 5" to 35" from horizontal. This variable slope permits controlled flow of material through the machine. Rubber paddles rotate inside the cylinder and are secured to a center shaft 2 in. square, which extends through the entire length and is turned at both ends so that it can be supported by sleeve bearings centrally fixed in the end plates of the shell. At the upper or feed end it extends through the bearing on the end plate into a sheave, to which the motor drive is connected by means of Texrope transmission. The four sets of paddles mounted on the shaft, two in each set, are made of rubber belting, 1/4 in. thick, 9 in. wide, and 16 in. long. The two paddles in each set are secured on opposite sides of the square shaft, but in the same longitudinal position along the shaft axis. In other words, the two paddles in the set operate 180" apart. Each successive set is mounted 90" from the preceding one. One end of each paddle is bolted to the square shaft and the other weighted by a steel bar, lx1 1hx9 in. long, which is attached about Yz in. from the outer tip of the paddle. By centrifugal force at proper shaft speed, these weights force the paddle surface to rub smoothly against the rubber lining of the cylinder. During operation, the sand forces its way between the paddle face and the surface of the cylinder, providing the necessary abrading action as the pulp flows by gravity down the length of the cylinder. The capacity of the machine is limited by the amount and the hardness of the coating material and by the quality of product desired. The degree of scrubbing is governed by the retention time in the scrubber, which in turn may be changed by varying the feed rate, water content of the feed, and

Jan 1, 1953

By M. G. Benz, J. F. Elliott

A new type of solution calorimeter has been constructed to measure heats of mixing, enthalpy increments, and heats of fusion, formation, and reaction at temperatures above 1000°C. With it, measurements hove been made of the partial molar heats of mixing joy tin in the Cu-Sn system from 0 to 13 at. pet Sn at 1127°C and fir nickel in the Cu-Ni system from 0 to 9 at, pet Ni at 1200°C. The integral molar heat of mixing for the Cu-Sn system is strongly negative with a calculated minimum value corresponding to the composition Cu3Sn. The integrnl molar heat of mixing for the Czi-Ni system is slightly positive at these temperatures. These integral molar values have been measured with a probable error less than 30 cal per g-atom at an atom fraction of nickel or tin equal to 0.1. The estimated probable error for the Cu-Sn system is approximately 450 cal for a measured heat of solution of approximately 9000 cal. For the Cu-Ni system it is approximately 700 cal for a measured heat of' solution of approximately 60 cal. A new type of heat of solution calorimeter was designed and constructed to provide a more accurate experimental method for determining heat effects at high temperatures, (T > 1000°C). With it, partial molar heats of mixing, enthalpy increments, heats of fusion, and heats of formation and reaction for metal systems can be measured directly. Previously it has been necessary to rely upon the use of the dropping calorimeter or the van't Hoff equation applied to equilibrium data for this information. For the most part equilibrium data are scarce and difficult to obtain in many cases, and it has been necessary to rely upon the "method of mixtures" for the desired enthalpy measurements. An earlier study1 provides a prototype for this work. The "method of mixtures" as outlined by K. K. kelley,2 "consists of dropping the substance under investigation from a furnace, with controlled and measured temperature, into a calorimeter operating at or near room temperature.'' In this manner the change in the total heat content (enthalpy) of the substance between the temperature of the furnace and that of the calorimeter is determined. In order to determine high-temperature heats of mixing, enthalpy increments, and heats of fusion, formation, and reaction, it is necessary to take the difference between two such measurements, one before and one after the occurrence of the enthalpy change being studied. The desired enthalpy change is then the small difference between two large numbers. Small errors in the large numbers are included as large errors in this small difference. It is this difficulty that must be avoided for accurate measurement of the desired functions. The procedure with the new solution calorimeter consists of dropping the substance under investigation from a furnace, with controlled and measured temperature, into a solution calorimeter operating in the temperature range of 1000o to 1200°C. Measurements in the range of 1200" to 1600°C are planned for the future. By selection of the composition and

Jan 1, 1964

By Gordon W. Hodgson, Edward Tipman

INT'RODUCTION An appreciable number of the oil fields in Western Canada are accumulations of heavy black oils in more or less unconsolidated sandstones. When the crude oils are produced by conventional means, they are frequently contaminated with both sand and oilfield water. The water is usually present as emulsified water, and the sand and other solid materials are suspended in the emulsions. The presence of the solid material in the crude oils is troublesome and, in general, little is known about the behavior of suspended solid particles in emulsions. The Stokes equation describes the behavior of suspended particles in homogeneous fluids in terms of particle size, density differential, and fluid viscosity. In the case of an emulsion, particle size and density differential introduce no difficulties, but considerable doubt centers around the viscosity term. It has been reported1 that the viscosity of an emulsion is related to the viscosity of the continuous phase by an expression of the type: n = a, (1 + a C) where n is the viscosity of the emulsion, n,, is the viscosity of the continuous phase, a is a constant, C' is the volume concentration of the disperse phase. The value of 17, the coefficient of the concentration term, has been determined for various systems, and has been found to vary from a value of 1 to 4.75. The value of 2.5 has been found to be applicable to a large number of systems, and this is the value that is widely used in the equation which then becomes the Einstein equation for the viscosity of emuls~ons. It is of considerable significance that neither the particle size nor the viscosity of the disperse phase enters the expression; the expression for the viscosity of the enlulsion is based solely upon the viscosity of the continuous phase and the concentration of the disperse phase. For crude oil emulsions the Simha equation and the data presented by Fukushima and lchimurai show a non-linear dependence of the viscosity ratio on the water concentration. While it is undoubtedly reasonable to use these data for predicting the viscosity of a specific emulsion, and while it is equally reasonable to expect the calculated viscosity to predict the settling rate of a solid spherical particle in the emulsion by means of Stokes equation, there seems to be no direct evidence that this calculation is indeed sound for describing the settling rates of solid particles in emulsions of water in crude oils. PROCEDURE A direct approach to the problem was thus indicated. Laboratory equipment was designed and built for the direct determination of the rate-of-fall characteristics of solid particles in emulsions of water in oil. Although the apparatus incorporated the basic design of conventional falling-ball viscometers, the study was concerned directly with the rate-of-fall characteristics of particles in emulsions, rather than with the viscosity of the emulsions as such. The crude oil used in this study was a very heavy black crude oil—the McMurray oil obtained from oil-soaked sands outcropping on the Athabasca River in northeastern Alberta. It was produced from the sands by strip mining followed by hot-water separation of the oil in the separation plant of the Alberta Government at Bitumount'. The primary plant product was a wet crude oil containing about 5 per cent solids and 35 per cent water. The final plant product was only slightly different from the native oil, having lost only about 4 per cent of its bulk as light ends during dehydration distillations. In the present investigation two kerosene dilutions of the McMurray oil were used in addition to the dry plant product. The McMurray oil contained 1.92 per cent solids. Analysis of the solids by X-ray diffraction showed the presence of kaolinite, pyrite, and illite as major constituents. with lesser amounts of quartz. Particle size of the pyrite and quartz was less than 10

Jan 1, 1957

By D. F. Stoneburner

The thermoelectric power and electrical conductivity of a series of liquid alloys of thallium were determined in order to study the relation of chemical bonding to semiconduction in liquids. The results indicate metallic behavior for the alloys of thallium with arsenic, antimony, and bismuth, while the sulfides, selenides, and tellurides indicate semiconduction. The results are compared with the Properties of the same alloys in the solid state. This work supports the observation that long-range order is not a necessary requirement for semiconduction. PREVIOUS work1,5 has shown that some sulfides and oxides behave as semiconductors in the liquid state. Although the band theory has been of great help in understanding the properties of solid semiconductors, its quantitative application to liquids is difficult. The very existence of liquid electronic semiconductors shows that ascribing the discrete structure of the energy spectrum to strict lattice periodicity, or long-range order, is not always satisfactory. Rather, the short-range order determines the structure of energy levels and the width of the forbidden gap and, consequently, the concentration of free holes and electrons. This is borne out by the analysis of data accumulated by Regel and coworkers.6 The similarity of electronic structure of liquid and solid phases of certain liquid metals was pointed out by Knight et a1.,7 who examined nuclear resonance in solid and liquid Li, Na, Rb, Cs, Mg, Al, Ga, Sn, and Bi. They found noticeable changes in the electronic structure upon melting only in the cases of bismuth and gallium. They concluded that "a liquid metal possesses a band structure which is very like that of the solid, provided the short-range structure is not altered during melting." In semiconductors in which the near-neighbor arrangement remains the same on fusion, the material would be expected to retain its semiconducting properties. The quantum theory of semiconductors, with its band structure of energy levels, is linked exclusively to long-range order. Furthermore, it has limited usefulness in predicting new semiconductors. Many sulfide8-12 have made comparisons of properties on the basis of crystal structure or the position in the periodic table, both methods involving the same factors as chemical bonding. Mooser and Pearson8"10 have pointed out that crystal structure depends on chemical bonding, and that this bonding is a property of the atoms themselves. As the periodic arrangement of the atoms offers an easy comparison of such factors as atomic weight and number of valence electrons it provides clues to expected similarities of bonding for various elements or groups of elements. This approach has led Mooser and Pearson to formulate their valence-bond theory of semiconduction, in which they define a semiconducting bond. These authors state that "semiconductivity is the result of the presence in a solid of predominantly covalent bonds. These lead, through the process of electron sharing, to completely filled s and p orbi-tals in the valence shells of all atoms of elemental semiconductors while in semiconducting compounds only one atom of any two bonded together need acquire closed s and p orbitals. The presence of empty 'metallic' orbitals in some atoms of a compound does not destroy the semiconductivity provided that these atoms are not bonded together; it may, however, give rise to pivoting resonance in

Jan 1, 1965

By W. E. Showalter

This paper discusses some of the results of combstion-drive tests which were made in a test cell using a sand bed 10 in. in diameter x 10-ft long. The test method is illustrated and described. The relationship between the API gravity of the in situ oil and the amount of air required for combustion drive is discussed in detail. Other things constant, the air requirement for combustion drive increases as the APl gravity of the in situ oil decreases. If the test results apply to actual reservoirs, the lowest-priced oils may cost the most to recover by this method. Information is shown which indicates that the effect of pressure on the amount of hydrocarbon burned is not large. A method of predicting air requirements from the API gravity of the in situ oil is presented. INTRODUCTION Combustion drive is the term used to identify the process of interstitial or in situ burning as an oil recovery method. Part of the in situ oil is burned to generate the energy needed to produce the remainder of the oil. Combustion drive as an oil recovery mechanism remains an economic uncertainty in spite of all the work that has been done by the industry in both laboratory and field?-6 This paper will show some of the results of tests which were made in a test cell for the purpose of studying the nature of the combustion-drive process. It will present data which indicate that the API gravity of the in situ oil is a significant indicator of the amount of air required to drive a burning front through oil sand. Air requirement varies inversely as the API gravity of the in situ oil. EXPERIMENTAL The tests were performed in a cell which utilized a cylindrical sand section 10 in. in diameter x 10-ft long. The thin-walled metal pipe which held the sand was wound with twenty external electrical resistance heaters which, by means of an automatic controller, maintained adjacent sections of the wall of the pipe at temperatures equal to the temperatures of the contained sand. Each heater covered 6 linear in. of the pipe. By this means lateral heat loss from the sand section was minimized, thereby causing the sand section to simulate more closely a horizontal increment of a combustion-drive reservoir. Fig. 1 shows a schematic diagram of the test assembly. Thermocouples to measure the temperature in the sand were located every 6 in. along the length of the sand section. The pipe containing the sand was enclosed in a cell designed for an operating pressure of 500 psig. The inlet air pressure was controlled at the inlet, and the gas flow rate was controlled and measured at the outlet of the cell. The oil sand used for the tests was prepared by mixing first water and then oil with the noncon-solidated sand using a closed mixer similar to a cement mixer. Table 1 shows a screen analysis of the sand. Ninety percent of the sand was 100 mesh or finer. This sand was a mixture of 80 per cent No. 120 Nevada White Sand and 20 per cent Tennessee Hi-Fusion Moulding Sand No. 3. The Nevada sand was a clean silica sand. The Tennessee sand con-

By W. D. Robertson

SINCE the comprehensive paper of Moore, Beckin-sale, and Mallinson,' little consideration has been given to the mechanism of mercury stress cracking of copper-base alloys, apart from extensive work on metallurgical and fabrication variables employing mercurous nitrate as an evaluation test. However, with the phase diagrams now available and the recent considerations of interfacial tension at grain boundaries, developed by Smith,2 it appears that the elements of a detailed solution to the problem are available. The facts that require explanation are: 1—Pure copper is not embrittled by mercury, with or without an applied or residual stress. 2—Cu-Zn and other copper-base alloys, in excess of some undefined solute concentration, are embrittled provided an external or residual stress is acting; cracking is not observed in the absence of stress. 3—The path of fracture is invariably intergranu-lar. 4—Copper and copper-base alloys do not dissolve readily in mercury at room temperature.' Consideration of the Cu-Hg phase diagram% hows that, at room temperature, solution of copper in mercury is probably limited by the formation on the surface of a microscopically thin layer of inter-metallic compound, CuHg, having a very limited range of solubility which effectively restricts diffusion of copper and mercury. But, above 150°C (the highest peritectic temperature) copper solid solution is in equilibrium with a liquid Hg-Cu solution and, consequently, copper should dissolve freely in a large volume of mercury with the formation of an equilibrium dihedral angle at grain boundaries in contact with the liquid phase. Depending on the magnitude of this angle, mercury will or will not penetrate the solid phase along grain boundaries and spread over grain faces. Thus the measurement of the dihedral angle of grain boundaries in equilibrium with the liquid phase above 150°C should indicate whether copper is intrinsically resistant to embrittlement by mercury. In view of the fact that mercury is appreciably soluble in copper, it may also be anticipated that diffusion of mercury will take place at higher temperatures in the absence of the intermetallic compound and possibly at a more rapid rate along grain boundaries than through the grains, resulting in a Cu-Hg alloy of unknown properties. In the case of brass, the ternary system Cu-Zn-Hg is involved and, from the resistance of brass to general solution in mercury at room temperature, it is probable that a compound of limited solubility range also exists, similar to the binary Cu-Hg sys- tem. Also, compared with copper, a change in the dihedral angle at grain boundaries in contact with mercury may be expected to result from the addition of zinc to copper. The preceding generalizations appear to be confirmed by the following experiments. Experimental Results Oxygen-free copper wire, annealed at 700°C for 3 hr in copper foil, was immersed in mercury at 170°C, with and without an applied stress. After 24 hr immersion, specimens which were spring loaded to a nominal stress of 10,000 psi had not failed, and there was no apparent embrittlement as indicated by a 180" bend around a diameter approximately equal to the diameter of the wire (0.125 in.) : General solution of the surface had taken place, indicated by a change in diameter of the wire, and the resulting structure of the surface is shown in Fig. 1. It is apparent that angles have developed at grain boundaries and examination at a high magnification indicates that the angle is in excess of 120°, completely excluding penetration of mercury. In view of the measurements made by Smith n the

Jan 1, 1952