US2285837A - Method of producing the abrasive metal carbides - Google Patents

Description

June 9, 1942. R, R RlDGWAY 2,285,837

METHOD OF PRODUCING THE ABRASIVE METAL CARBIDES Filed Nov. 23. 1940 lll/l1 /l/ Patented .une 9, 1942 UNITE il? STA PTENT- orrica Mn'rnon or raonucnvo 'mn anaslva METAL caaamns 7 Claims.4

This invention relates to methods'of producing the abrasive metal carbides and more particular ly the carbides of boron, silicon, tungsten, titanium and zirconium.

These abrasive metal carbides are sensitive to the influence of certain gases, such as oxygen and nitrogen, at temperatures at, whichthey are formedand in some cases at materially lower temperatures. For example, boron and titanium carbides are attacked by nitrogen and oxygen not only at the temperature of their formation but also during the initial'portion ofI the cooling stage. 'I'hey also dissolve carbon when in the molten condition. Because of this sensitivity to substances normally found in furnace environments, special means has been required'heretofore for producing these carbides on a large scale without serious contamination.

'Ihe standard procedure for making silicon carbide from a mixture of sand and coke has involved the use of an open walled resistance furnace, in which the reagents are enclosed in a very large volume of reducing material which prevents access of oxygen and nitrogen lto the charge. But this requires providing and handling large masses of material to produce a much smaller amount of final product. Because of the high cost of some of the other raw materials, it has notl been practicable to use them liberally and on a large scale as an enveloping medium or sealing container. Furthermore, these enveloping materials have permitted only a slow cooling of a molten abrasive metal carbide and gases have been able to seep through the seal and contaminate the'product. The carbides of boron,

titanium and tungsten absorb nitrogen readily when they are in a hot condition, and there is no protection afforded by the enveloping mixture, since the atmospheric gases work their way through the outer charge and by reaction with the carbon thereof produce carbon monoxide and nitrogen and the latter contaminates the metal carbide. The presence of any material amount of a nitride of the metal alters the abrasive char.

acteristics of the material to a very serious extent.

Consequently, boron and titanium carbides have been made in a closed furnace and in accordance with the methods and apparatus disclosed in my U. S. Patents No. 2,123,158, of July 5, 1938, land No. 2,155,682, of April 25, 1939. These furnaces are closed, cylindrical structures having small gas egress openings, in which the high temperature required for the electrical synthesis is provided by a resistor embedded in the charge. The entire charge is packed into the furnace casing around the graphite resistor, and

. the electric current is passed through the resistor lilo 'within this entire charge until either the resistorY has been consumed or broken and the current interrupted or until enough of the charge has been converted to form a. large button or ingot of the nal product. Then the furnace is dismantled and the partially converted and raw materials are removed and separated. The unconverted materials are then available for another furnace charge but only after their'chemicai character has been ascertained by analysis and the proper stoichiometric calculations have been made. A product of high .purity may be thus obtained.

Onl the other hand. there are certain inherent advantages in the electric arc furnace. lFor example, the furnace may be comparativly small and yet operated on a large scale, since the reaction may be carried on continuously .by the gradual feeding of a properly proportioned charge to the high temperature zone of the arc and thus progressively building up a large mass of reaction product. The power may be readily adjusted to the desired scale of operations and this provides great flexibility-in plant procedure. The operation of the furnaceis simple, and its control effective to produce a desired composition and structure.

'I'he abrasive metal carbides are made by heating the metal oxide withcarbon proportioned stoichiometrically for the desired composition.

The metal carbide is very refractory' and requires temperatures upwards of 2500 C. for its production and ultimateV crystallization. This extremely high temperatureis only achieved on a moderate scale by the use of a. large quantity of electrical energy concentrated into a very small volume of reactivematerials. The s'implest way of obtaining that large concentration of power in a small `space isby means of a submerged axc. The electric arc has been used satisfaotorily in the production of crystalline aiumina by melting the non-crystalline material in an open top furnace of the type shown in the patent to Higgins, No. 916,866, of March 30, 1909. It, however, is characteristic of the submerged arc, open top furnace that the action'oi. the electric are is such that -the enveloping furnace charge is punctured and the furnace gases flow in eddying currents around the electrodes and over the furnace wall. At the same time, the

non-uniformgas flow permits air currents to enter the furnace and particularly at the sides in a line at right angles to the plane of. the electrodes. TheA air diffuses into the cell and is drawn into the reaction zone where it attacks the flnal product and leaves nitrides and suboxides therein. The components of the air also unbalance the reaction by attacking the metal vapors, and causing themto escape from the cell largely as fumes of boron oxide or other metal oxide. Consequently, the electric arc furnace as heretofore constructed has not been availablefor the production vof these abrasive metal carbides, which are sensitive to oxygen and nitrogen, since it has been impossible to maintain a continuous seal against the entrance of atmospheric gases.

The primary object of this invention is to make these abrasive metal carbides in an open top electric arc furnace land to provide a method which will serve eiilciently for their production.

A further object of the invention is to provide a method of making the abrasive metal carbides progressively vby reacting small increments of charge in an electric arc and gradually producing a large mass of product under controlled conditions. t

A further object is to provide a method of producing rapidly and efficiently in an electric arc furnace an abrasive metal carbide of high purity and which is not materially contaminated with over the product at all times during the furnace run. Also, the product il cooled under similar favorable conditions. The furnace cell is, therefore, so constructedv as to insure that this procedure may be carried out, and this comprises a structure and arrangement of furnace parts wherein the gas outlet is'of such restricted size and location that the reaction gases will move continuously outwardly from the reaction zone throughout the entire exit area. 'Ihe exit may be of small size and preferably located at a point remote from the reaction zone. While the furnace walls may be variously shaped to provide these required conditions and to insure that the flow of gas is only outwardly,I I prefer to 'utilize a furnace cell shaped asa cylinder having an open top, and the above conditions are satised if the height of the cell is at least three and preferably four times its diameter. The proper ratio of these dimensions, the location of the opening and the shape of the furnace depend upon the particular reaction to be carried out dissolved carbon or oxygen and nitrogen reaction products. Further objects will be apparent in the following disclosure.

Referring to the drawing:

Fig. 1 is a cross sectional view of an electric arc open top furnace which will serve for the production of these abrasive metal carbides;

Fig. 2 is afragmentary sectional viewv of the eration; and y Fig. 3 is a sectional view of the prior art type of open top electric arc furnace used for the fusion and subsequent crystallization of alumina.

Referring rst to Fig. ,3 of the drawing, I have there shown a somewhat diagrammatic view of the Higgins arc type furnace used for making crystalline alumina. This comprises a metal container i0 carrying 'a' carbon bottom H. Two or more electrodes I2 are suitably suspended within the wide open top of the furnace and the bottom portion of the furnace prior to its opcharge of alumina is fed intermittently through the opening as required for the melting operation. If a furnace of this type were used normally for the production of boron carbide or titanium carbide, for example, the general flow of the gas currents would be somewhat as indicated by the arrows, and wherein the suction of air intov the furnace would be maximum at right angles to the plane of the electrodes. That is.. the evolution of carbon monoxide travels upwardly along the electrodes and largely between them and this produces a reverse current flow of air over the sides of the furnace and down into'the high temperature zone where the oxygen and nitrogen are free to attack the reagent metal vapors and the carbide that is being formed.

I have discovered that these abrasive metal r carbides may be made in the electric aro type of'furnace provided the furnace is so operated that the gases produced during the reaction are sumcient to keep lled all of the free cell space and they are caused to ow outwardly with such a velocity and direction of travel that the atmospheric gaseawlll not enter the reaction zone as eddy currents or by any back diifusion. The power input to the furnace is so controlled as to produce a rapid evolution and a large volume of gas and maintain it as an inert atmosphere and they are controlled especially by the size of the equipment and the power rating of the furnace; .hence there may be considerable variation inthese characteristics.

The abrasive metal carbides may be efficiently made ina furnace of the general shape and construction shown in Fig. 1 of the drawing. This' furnace may comprise a cylindrical iron shell 20 having a closed iron bottom 2| and an open top. The shell may be cooled by a water pipe 22 encircling the top which is so arranged as to supplya constant stream of cooling water to the outer surface of the shell. This results in the formation ofl a crust of unconverted or partially converted materials on the inner face of the shell which serve as a refractory lining and protect the shell from the heat. The container may be provided with a bottom 24V of carbon, such as graphite block, which serves as a heat insulating medium. The electric arc is preferably developed by means of two graphite electrodes 25 suitably connected to an alternating current power circuit for the introduction of a high amperage low voltage current. The electrodes are long and slender and proportioned to carry the load. They may be made of a plurality of cylindrical graphite sticks secured end to end by screw threaded-plugs threaded into' sockets in the adjacent ends of the electrodes. By securing together these short lengths to form a long continuous conductor, I provide an elongated electrode of small cross section which will reach to the bottom of a very deep furnace cell without having a large amount of power wasted in the electrode which would cause its early destruction. The electrodes are suitably suspended for vertical movement by means of water cooled clamps 2l secured to their top ends and which are carried by suitable metal supports 28 arranged to be moved vertically between guides 29. The electric power may be introduced through the flexible cables 30 suitably connected to the clamps. Various ltypes of apparatus well known in the electric furnace industry may'be employed in this connection, and suitable control apparatus may be used to raise the electrodes and insure constant current conditions as required. The electrodes are vertically moved as the charge is added, and-the final product is developed as a progressively enlarging mass located beneath the electrodes.

Asan example of the proportions which are satisfactory for operating an arc furnace withtop out nitrogen or oxygen contamination .of the product, I may make titanium carbide in a cylindrical furnace cell having a diameter of 12 inches `and a height of 48 inches. The electrodes are graphite electrodes 3 inches in diameter. For the production of titanium carbide, a power of from 75 to' 125 kilowatts may be employed with a voltage of 50 to 80 volts.

In order to prevent carbon contamination of a molten product, such asB4C or TiC, while the molten material lies in a pool at the bottom of the furnace, I preferably place a deep layer 32 'of coarse particles of titanium oxide on the carbon bottom, and this coarse layer may, if desired, be itself supported ona layer 33 of nner material which further aids in holding the pool of molten material out of contact with the carbon bottom. The only other carbon present, which is found in the electrodes and the charge, cannot contaminate the molten mass under the operating conditions of the furnace.

In order to start the operation, some ofthe carbon, which is required for reduction ofthe titania of the furnace charge, may be kept out of the charge mixture, and this is segregated into a small trench laid directly between the electrodes. This trench material 34, as indicated in Fig. 2, is so arranged that the electrodes 25 may be brought down into contact -with the carbon and an electric arc struck across this material, after which the electrodes are gradually raised to control the current iiow. This conducting The melting furnace conditions may also be suitably regulated to produce a material of required oxide content. That is, the furnace may be operated under reducing conditions to provide the lower oxide TizOs, or'I `may use one which is chiey made up of TiOz. When the substantially minimum reducing conditions are employed in the Higgins type open arc furnace and the material is subjected to standard 'furnacing condition's, some reduction may take place due to the use of the graphite electrodes; but if greater reducing conditions are maintained, then a material may be produced which contains from to 30% of TizOa with the remainder made up material in the trench is preferably made of graphite electrode material that has been crushed to a small particle size. charge mixture is started so as to cover the incandescent arc and to generate the reducing gases which blow out of the top of the elongated furnace shell. The power input is regulated to provide a high power rate and the feed of the charge is rapid enough to maintain the blanket covering over the arc. This insures a high rate of developing CO gas and a high velocity of gas movement to and through the cell opening at the It has been found desirable to use coarse particles of the metal oxides mixed with coarse granules of coke in order to permit a smooth opere ation of this furnace, which represents a very high power input per unit-volume. Sawdust is preferably mixed with the charge to aid in keeping it porous. These coarse` granules permit the free and even escape of gases and minimize they disturbance of both the charge and the metal carbide products produced by the high temperature reaction.

The high purity titania available in commerce is a chemically precipitated titanium dioxide which has a very low density and is a light fluify material. Its apparent density is about 0.7 gram per cc. In accordance with the procedure set forth in my prior application Serial No. 137,620, iiled on April I9, 1937, I fuse this uify titanium oxide in a Higgins type open arc furnace ofthe general construction shown in Fig. 3 of the drawing, or in any other suitable furnace, and then crystallize it as a material of high density. 'The titania powder is charged slowly into the furnace where it is electrically heated to a temperature above 1600 C. and melted by the arc struck between the graphite electrodes. After a suiiicient amount of material has been melted, the electrodes are withdrawn andthe mass is allowed to cool rapidly to form small crystals of a size which is suitable for the reduction stage.

The addition of the chiefly of Ti02, with small amounts of other suboxides present. This dense material after it has been fused and crystallized has an apparent density of 2.44 grams per cc. and a real density of 4.26. YThe chemically precipitated titania which has a purity of 99.5% is further purified during the furnacing operation with a loss of silica and sulfate so that the final product has a high degree of purity.

This prefused titanium oxide of high density, as made in accordance with the above described procedure, is crushed to a desired size, such as one which will pass through a screen of 4 or more meshes tothe linear inch and it is then mixed with the carbon reducing agent and preferably a pure carbon, such as petroleum or pitch coke or high quality which has a low ash content. The carbon reducing agent may also be -made up oi' particles of comparatively large size,

such as those Whichwill pass through a screen of four or more meshes per linear inch, and the titania and carbon are mixed together to provide a substantially uniform mixture. This mixture is proportioned stoichiometrically according to an analysis of the oxide content of the titania and the chemical composition of the carbide product which may be TiC or a titanium rich TiC. The carbon is proportioned for removing the oxygen as CO and for combining with titanium as TiC. This charge mixture is fed continuously or periodically to the furnace as above explained to maintain a refractory heat insulating porous cover over the electric arc and the molten pool there below,l and this rate of feed is such that the cover is ordinarily kept thin to prevent reaction with the electrodes at this point.

The layer of l pure titanium oxide underlying the arc keeps the molten titanium carbide from contacting with the carbon therebeneath and so prevents contamination by carbon which is readily soluble therein. If the fluffy chemically precipitated titanium oxide powder were used instead of the crystallized material above described. there would be a serious tendency for this powder to blow around in the violent gas currents generated Within the furnace chamber and in fact to blow out from the top of the cell. This turbulent action would also prevent. the formation of a quiet pool in the bottom of the cell and it would readily stir up the bottom layer and penetrate to the graphite where it would become contaminated. The relatively dense and heavy crystals of titanium oxide on the bottom of the cell are not disrupted or disturbed by the gas currents and the coarse particles of the charge are likewise heavy and dense and do not become disturbed seriously by the outfiowing gas. Hence the action proceeds smoothly and Without material disturbance of the pool of liquid forming beneath the electrodes. Y

The volume of the reaction gas may be partly controlled by adjustment of the power input. The higher the power rate, the 'more rapid is the gas evolution. I'he power input is to be sufficient to maintain a desired reaction temperature, such as 2600 C. or higher. I may also increase the volume and velocity of the outpouring gas by adding kerosene or other hydrocarbon oil to the mixture. This permits a lower power rate to insure the presence of enough gas and vapors to iill the cell space and sweep rapidly outwardly any air that tends to get into the furnace.

Owing to the narrowness of the furnace cell or of its opening, the conilned gases are forced to pour outwardly from the cell with a fairly violent action or high velocity and thus sweep along with them any air which might tend to eddy into the cell over the top wall thereof. It is, therefore, important that the cell confine the reaction zone or the cell outlet to such an extent as to cause this upward rush of liberated gases at a vcomparatively high velocity. This method, therefore, involves subjecting the charge to the high temperature conditions of an arc located above a pool of the final product and wherein the gases are produced at a rapid rate and are forced by the conflning walls to travel outwardly at a high velocity and thereby prevent ingress of atmospheric gases. This insures that the final product is always bathedvwith carbon monoxide or a non-oxidizing gas and that no nitrogen can contact with it to form a nitride.

Although the drawing shows only two electrodes, it will be appreciated that a multi-phase current may be used, such as a three phase current, and that three or a larger number of electrodesmay be accordingly employed. I prefer to use the two electrode system, since the presence of more than two electrodes tends to interfere with the outward smooth flow of the produced gases. The electrodes are periodically or progressively raised as required to maintain proper fumacing conditions, anda quiescent pool of molten material grows gradually on the titania layer 32. The molten mass is ultimately chilled and crystallized by the water cooling of the shell. The electrodes may be raised and the ingot thus progressively increased in size until there is dan- 1 ger of contamination vby the oxygen and nitrdgen of the atmosphere and then the furnace run may be considered as completed. The electrodes are then removed and the mass may be allowed to stand in the shell under natural or controlled water cooling conditions for a suiliclent length of time to insure that the ingot has solidied. The rate of crystallization or solidiilcation may be controlled by controllingthe water how. Various expedients well known in the industry may be adopted for producing the desired crystalline structure. v

Since titanium carbide, as well as boron carbide and other products tends to become oxidized or to react with nitrogen during the cooling stage, I also take steps to insurel that the product may .cool to a safe temperature and without danger of the atmospheric gases inltrating to the product.

Because of the shape of the furnace and its small Y size in relation to its power input, the product cools rapidly as soon as electrical energy is shut v ofi, hence it is ordinarily suillcient merely to cover the product with a thick blanket layer of the charge which is rich in carbonaceous material. I may also provide a cover oran insert gaseous oil, such as kerosene oiL'ontQf-the-top of the hot furnace charge and allowing the vapor products e material by steadily dripping alight hydrocarbon assess? thus formed to keep the furnace chamber full of gases which steadily sweep out any atmospheric air that tends to diffuse into the cell chamber.

Titanium carbide of the formula TiC, as manufactured in accordance with this method, is of high purity which may run above 99.8% of TiC. It has a high metallic luster and well developed crystal faces and sharp edges, and it is a dense well crystallized structure. Its density is greater than 4.75. The crystals form in a massive hexagonal columnar structure, and they may be made to crystallize in a grit size larger than will pass through a screen of meshes per' linear inch. It desired, the sinichiometric proportions of the furnace charge may be varied and the power input regulated to produce a product having a high' titanium content in which the productI is made up of titanium carbide presumably cemented together by titanium metal. The product does not contain titanium oxide or partially reduced titanium carbide.

It will now be appreciated that the other abrasive metal carbides may be made in accordance with the above procedure. not fuse at the temperature of reaction but forms directly in its crystalline state, but otherwise the procedure may be carried on as above indicated. The charge may be made stoichiometrically of silica of a suitable grade with coke, such as petroleum coke, proportioned in accordance with the reaction SiO2+3C,=SiC-i-2CO. Boron carbide, titanium carbide and tungsten carbide are extremely sensitive to oxygen and nitrogen, and the above precautions speciiied in regard to titanium carbide are desirable in order to prevent the formation of the oxides or nitrides of the metals. The chargefor each of these materials will be proportioned stoichiometrically to form the desired carbide and carbon monoxide as above indicated.

For making boron carbide I may employ the procedure of my prior Patent/No. 1,893,106 of January 3, 1933, which serves for producing a boron oxide glass that' may be mixed with carbon in coarse granular condition in accordance with the stoichiometric proportions represented by the formula 2BaOa+7C=B4C+6CO. Moreover, the procedure of my prior Patents No. 1,897,214, of February 14, 1933, and No. 2,141,617, of December 27, 1938, may be adopted in connection with the production of boron carbide except as changes are required in the apparatus and process in accordance with the above described method of employing an electric arc for the heating medium.

The same general principles apply in the production of zirconium and tungsten carbides and the a be understood that this invention covers the production of the carbides of these metals whether or not another phase, such as the metal, is present therewith. That is,"'I maymake a product which is an alloy or a composition of matter having a poly-phase of crystalline boron carbide of the formula BiC intimately associated with or cemented together by one or more phases of a boron rich material which has boron for its primary oonstituent and is substantially free from. uncombined carbon in the form of graphite. Similarly, a carbide product rich in titanium, tungsten or zirconium may be made in accordance with the procedure of that patent. This method and the furnace construction have many advantages. The procedure is characterized by its simplicity of operation and by its exibility as regards the Silicon carbide does l quantity of charge to be reacted. The furnace dimensions and shape may be as desired within the principles of this invention, and one may adjust the power input in accordance with the scale of the furnace. It is possible to obtain a continuous feed of material and it is, therefore, not necessaryto place all of the charge in the furnace at one time. 'Ihe operation of the furnace may be regulated by feeding the charge as needed to maintain the enveloping heat refractory blanket and to permit the reaction to go on undera controlled rate.

It will now be appreciated that many modi- :ations may be made in the furnace structure and hat the procedure of making the carbides may be varied within the knowledge of one skilled in the art. Hence the above disclosure is to be interpreted as describing the principles of the invention and my preferred process and apparatus and not as limitations on the claims appended hereto.

I claim:

1. The method of making boron carbide comprising the steps of mixing granular carbon and coarse granules of dense boron oxide in stoichiometric proportions for making boron carbide ofdesired boron content, developing an electric arc in a deep open Iended cell and maintaining the reaction temperature, feeding the charge mixture gradually to the arc and forming molten boron carbide which crystallizes as a growing ingot, maintaining a substantially unobstructed cell space above the reaction zone which is several times deeper than the width of the cell opening, controlling the power input to produce reaction gases rapidly and causing them to travel continuously throughout said cell space and prevent the ingress of atmospheric gases to the reaction zone, maintaining the molten boron carbide out of contact with excess soluble carbon during both the forming and cooling stages, and cooling the boron carbide within the cell in an inert atmosphere devoid of oxygen and nitrogen to a temperature at which it is stable in air.

2. The method of making titanium carbide in an electric arc furnace cell having an open top comprising the steps of mixing stoichiometric proportions of coarse granular carbon and coarse granules of crystalline titania, feeding the charge mixture gradually to an electric arc within the cell while controlling the electric current to maintain a rapid reaction and progressively forming and crystallizing as a growing ingot a pool of molten titanium carbide while preventing the molten material from contacting with excess carbon, maintaining a substantially unobstructed cell space above the reaction zone which is several times deeper than the width of the cell opening, causing the reaction gases to iiow continuously and rapidly away from the reaction zone and to sweep from the cell top throughout the entire free area thereof and prevent the ingress of the atmosphere, and cooling the carbide within the cell in an inert atmosphere free from oxygen and nitrogen.

3. The method of making tungsten carbide :omprising the steps of mixing stoichiometric )roportions of granular carbon and granular .ungsten oxide, feeding the charge progressively o an electric arc in a furnace cell having an open op, controlling the power input to provide the reaction temperature and form a gradualy enlarging mass of tungsten carbide, maintaining a substantially unobstructed cell space above the Y reaction zone which is several times deeper than the width of the cell opening, ,causing a continuous rapid movement of'the reaction gases away from the reaction zone which sweeps from the cell outlet all atmospheric gases and prevents ingress thereof to the final product, and cooling the carbide within the cell in an inert atmosphere devoid of oxygen and nitrogen.

4. The method of making an abrasive metal carbide comprising the steps of forming an electric arc adjacent to the closed bottom of an open topped, deep furnace cell, progressively feeding to the arc through the open cell top a layer of a stoichiometrically proportioned charge of carbon and an oxide of the metal, controlling the' electrical power input and the cell temperature to build` an ingot of said carbide above the .cell bottom, maintaining during the furnace run an unobstructed cell space above the arc and the ingot which is several times deeper than the width of the cell opening and which permits the free egress of the gases, causing the evolutionof a large volume of confined gas which moves rapidly upwardly and seals thecell opening against the admission of atmospheric gases, and progressively raising the arc as the ingot builds up.

5. The method of making an abrasive metal carbide comprising the steps of forming an electric arc adjacent to the closed bottom of an open topped cell which is several times deeper than the width of its opening, progressively feeding a stoichiometrically proportioned charge of carbon and an oxide of the metal through the cell opening and forming a blanket thereof over the arc reaction zone while leaving the cell space thereabove substantially unobstructed by solid material, controlling the electrical power input and the cell temperatures to provide a high reaction rate and progressively build an ingot of said carbide above the cell bottom, while causing the evolution of a large volume of confined gas which moves rapidly upwardly through the unobstructed cell space and seals the cell opening and prevents the admission of atmospheric gases, and maintaining the reaction conditions by progressively raising the arc, and thereafter cooling the reaction product within the cell while maintaining a gaseous atmosphere thereabove which is substantially free from oxygen and nitrogen.

6. The method of making an abrasive metal carbide comprising the steps of providing an open topped cell whose depth is several times the width of the opening, continuously cooling the exterior of the cell, feeding progressively to an electric arc near the bottom of the cell space a stoichiometrically proportioned charge of carbon and an oxide of the metal without materially obstructing the cell space above the arc, controlling the electrical power input and causing the evolution of a large volume of gas which moves continuously upwardly throughout substantially the entire cell space and exit opening and prevents the admission of atmospheric gases to the reaction zone, and progressively developing an ingot of the carbide while raising the arcto maintain the reaction, and subsequently cooling the product within the cell while introducing a gas producing substance and maintaining with the cell an upward flow of an inert gas which is free from atmospheric gases.

7. The method of making an abrasive metal carbide in ar open topped electric arc furnace cell, comprising the steps of providing the bottom of the cell with a layer of the metal oxide on which the final product may rest, establishing an electric arc above said layer, progressively feeding to the arc a layer of a charge of carbon and dense particles of the metal oxide. cooling the exterior of the cell, controlling the electrical power input to maintain the reaction temperature and cause the progressive formation of a molten mass of the metal carbide which crystallizes` asa growing ingot, gradually raising the arc while maintaining a substantially unobstructed cell space thereabove which is several times deeper than the width of the cell opening, causing a large volume of gas to move continuously upwardly fromthe reaction zone and seal the cell opening against the entrance of atmospheric gases, and thereafter cooling the molten product in the cell while maintaining an inert. atmosphere thereabove which is devoid o1' oxygen and nitrogen.

' RAYMOND R. RIDGWAY.

Images