US3403009A - Refractory metal structures - Google Patents

Description

3,403,009 REFRACTORY METAL STRUCTURES Theodore R. Bergstrom, Little Canada, Minn., assignor to Minnesota Mining and Manufacturing Company, St. Paul, Minn., a corporation of Delaware N Drawing. Continuation-impart of application Ser. N 0. 388,614, Aug. 10, 1964. This application June 9, 1967, Ser. No. 644,820

6 Claims. (Cl. 29182.7)

ABSTRACT OF THE DISCLOSURE Articles made by sintering powdered refractory metals preformed into a desired shape with the aid of a filmforming, essentially non-volatile organic binder polymer which thermally degrades on heating to form carbon, sufficient amounts of said binder being present to coat the powder particles and to leave about 0.04 to 0.8 percent of carbon in the metal after firing. An unworked, sintered refractory metal article is thus formed, which has improved ductility and/or fine grain size. The fine grain size is retained even after the article has been reheated to say 2000 C.

This application is a continuation-in-part of my prior copending application Ser. No. 388,614, filed Aug. 10, 1964, now abandoned.

BACKGROUND OF THE INVENTION This invention relates to powder metallurgy and the production of metallic structures from powdered metals, and more particularly from powdered refractory metals.

The Group VI-B refractory metals as generally available are brittle at low temperature and must be worked and stress-relieved to produce mill shapes which are of practical utility. Further fabrication including welding, brazing, diffusion bonding, mechanical fastening, of formed sheet or forged, extruded or drawn parts is necessary to produce structures of practical utility. These operations are costly, time consuming and, particularly with respect to rolling, can be carried out only with simple, elongated structures or shapes such as sheets, angles, I-beams and the like. Often, repeated heating is necessary for fabricating even such simple shapes, as the more brittle metals can only be bent or drawn at high temperatures if cracking is to be avoided.

Furthermore, so far as it is known, it has heretofore been believed that in order to obtain these refractory metals in more ductile form, essentially all carbon must have been removed from the metal, and that ductile shapes are produced by mechanical working, with repeated heating. See, e.g. United States Patent No. 2,776,- 887; and Journal of Metals, February 1956, pages 150- 155.

, SUMMARY OF THE INVENTION According to the invention, it has been found that contrary to what was believed in the prior art, the presence of carbon in certain limited amounts in certain refractory metals makes possible the direct production of more ductile refractory structures, even of intricate form, without mechanical working, by powder metallurgy processes.

It is an object of the invention to provide a process whereby green structures can be made which on firing produce refractory metal structures of improved ductility in unworked form, which are suitable as intermediate forms for further mechanical working into sheet, bar, rod, tube and the like, and wherein the organic binder may function to reduce the oxygen content of the metal during sintering. Also contemplated is a process whereby nite States Patent refractory metals can be sintered to very high density at relatively low temperatures.

It is a further object of the invention to produce various forms of refractory metal structures which are unworked and of an exceedingly fine grain size, and which retain their fine grain structure after heating to very high temperatures, which can be formed as exceedingly complex structures of very high density without the use of pressure or mechanical working.

Another object is to provide a process for reducing the oxygen content of refractory metals during sintering to very low levels, and allowing the use of oxidized metal powders.

In accordance with the above, it has surprisingly been found that when small amounts of carbon, i.e. about 0.04 to 0.8 percent by weight, are uniformly distributed throughout a refractory metal (believed to be present as metal carbide and an actual solution of carbon in the metal, i.e. an alloy, as will appear hereinafter), the resulting metallic structure is relatively ductile and strong, even when mechanically unworked. The metal structures thus produced are without preferred orientation of the grains, but on examination of metallurgical samples under the microscope are found to have characteristic metallic grain structure in which metallic carbide is visibly present along the grain boundaries. The average grain size of the metal ranges from a few tenths of a micron up to about 25 microns.

It has also been found unexpectedly that the grain size in molybdenum, tungsten and other metallic structures produced by the methods disclosed herein remains ap preciably finer after exposure to temperatures in the range of about 18002000 C. than does that of structures produced by the conventional working processes.

By the term unworked as used herein, it is meant that the metallic article has not been pressed, rolled or otherwise mechanically worked during or after sintering. The green intermediate structures may, however, be pressed or rolled to some extent as a part of the green fabrication process.

The refractory metallic articles of the invention are produced by a process of powder metallurgy, in which a green structure adapted to firing is first prepared. This green structure comprises the refractory metal in finely divided form and a binder therefor, the binder being a film-forming, essentially non-volatile, organic polymer which thermally degrades on heating to form carbon. A sufiicient amount of the binder is employed to coat each particle of the refractory metal over substantially all of its surface.

These green structures can be made in the form of green sheets, which are produced with the aid of a solvent for the binder so as to produce initially a slurry, thick paste or plastic mass of metallic powder, binder, plasticizer and solvent. Even almost dry powder-binder aggregates can be used. This green mass is shaped by rolling, pressing, extruding, knife-coating, etc., evaporating the solvent if necessary, to form a desired shape. Dry, green shapes so formed are then temporarily adhered to form the configuration desired in the final structure, with proper allowance for shrinkage which occurs during sintering. This temporary adherence is readily brought about by cementing the shaped green articles, e.g. sheets together with a kind of glue made by adding solvent to a paste of metal and binder until the consistency desired is obtained, coating the contacting surfaces with this paste, and pressing them together. Alternatively, the surfaces to be joined can be softened by wetting with a suitable solvent for the binder and then pressed together. If a film of thermoplastic material is employed as a binder, heat seals can be made.

other forms has been made, it is carefully dried to remove A practically all of the solvent and pre-sintered to decompose the binder to make carbon available for oxide reduction, formation of carbides and/ or dispersion in solution in the metal grains, as an alloy. Any gaseous decomposition products escape through the partially sintered, somewhat porous mass.

The control of oxygen impurity is generally regarded as a central problem in improving the ductility of molybdenum and tungsten, and elaborate handling precautions have recently been advocated to reduce powder contamination during processing. Deoxidation using carbon is employed when the metal is made by fusion processes, i.e. arc casting or electron beam melting. The normal method of reducing the oxygen content of refractory metals in sintering operations involves the use of hydrogen or cracked ammonia gas. The use of carbon as a reducing agent in solid-state or sintering reactions is not generally practiced, because of the dispersion problem, the localized high carbon areas, and the reported detrimental effects of carbon on ductility.

Calculations from thermodynamic data show there are advantages in using carbon instead of hydrogen in reducing the equilibrium concentration of oxygen. However, the kinetic effects apparently are more profound. Thus, the removal of oxygen by hydrogen involves diffusion of hydrogen to the point of reaction and counter diffusion of the resulting H O, while removal by the carbon reaction involves only the outward diffusion of the resulting CO.

It is therefore believed that the organic binder material in addition to performing a bonding and configurational function, thoroughly coats all metal particles and upon thermal degradation to a carbon residue effectively reduces surface metal oxides by the evolution of carbon monoxide and produces the oxide-free surface on the particules that is needed for effective sintering. This de-oxidation will be quantitatively demonstrated in the examples that follow.

Some of the binder derived carbon will remain after sintering and will combine in the metal lattice to the extent of the solid solubility. Carbon in excess of the solid solubility content will be converted to refractory metal carbides during sintering.

The article thus produced is finally sintered to produce a dense, refractory metal structure corresponding in configuration to the original intermediate green structure. A certain amount of shrinkage takes place, depending upon powder size, green density and required density of the final structure.

By the present method it is possible to produce metallic articles having very high density, being of practically any shape and size, and having uniform density and ductility throughout. The parts of the articles are not Welded by the usual processes or soldered or brazed using different metals; instead the joints are sinter-welded and are of the same strength, density and ductility as the remainder of the article.

The metals which can be used in the process of the invention to produce the articles of the invention include molybdenum, tungsten and alloys thereof. For the process of the invention they are commonly employed in very finely divided form, i.e. 325 mesh or smaller, ranging down to sub-micron size. However, if desired, larger particles can be employed, and if special results are desired, variously sized metal powders can be used in combination.

The binders which are employed for making the green structures include those which will form films, which are organic in nature and which will decompose on heating to form carbon. To these may be added plasticizers to make the sheets which are produced more flexible, or solvents for reducing viscosity. Exemplary binders are methyl cellulose, ethyl cellulose, polyvinyl alcohol, cellulose acetate, phenolformaldehyde resins, urea=a'ldehydc resins, etc. If non-volatile plastiicizers are used, they will on decomposing when heated contribute some of the required carbon.

Suitable solvents include water, acetones, lower alcohols, fiuorinated solvents, etc. The exact nature of the solvent is ordinarily without significance, save that it should be practically inert toward the powdered metal which is employed, and preferably'is reasonably volatile. Aqueous solvents are preferred. v

When resins are used which polymerize in recognizable stages, as for example phenol-aldehyde resins or polyimides, they are usefully employed in their most soluble stages for preparing the green articles. Thereafter, polymerization may continue through the insoluble stages to form relatively rigid objects suitable for firing.

Exemplary plasticizers are such Well-known materials as glycerin, or synthetic waxes such as low-molecular weight polyalkylene oxides, etc., used when aqueous solvents are employed; other waxes, e.g. hydrocarbon or vegetable waxes, used when organic solvents are employed; and the like. It will be apparent that volatile plasticizers can be used in addition to such materials as may decompose to form carbon on heating.

The amount of binder employed is chosen to be that amount which will yield the desired quantity of carbon after firing. Empirical methods can be used to determine the best proportions of binder used for each set of conditions, including firing conditions; and thereafter, the same quantity of binder will yield the same amount of carbon under those conditions. If somewhat too much binder is used, the excess carbon can in many cases be removed by firing to higher temperatures, or by firing in wet hydrogen. If in particular instances this amount of binder does not give a useful flexible green sheet, a fugitive binder, for example poly-butene, polyvinylbutyl ether or the like, can be added in amount sufficient to render the film self-supporting. From about /2 to 10 percent or more of the decomposable binder can be employed, based upon the weight of the metal; preferably, /2 to 5 percent of binder is used.

It is common practice to produce articles from compacts made from metal powders using dies and presses. This type of processing limits the configurations that can be produced and requires expensive tools. Excessive pressures are required to achieve compacts that will sinter to desired final densities. It has been discovered that the above-mentioned disadvantages can be overcome by the process of the invention.

The following examples, in which all parts are by weight unless otherwise specified, will further and more specifically illustrate the preferred embodiments of the invention and the process for their production.

Example 1 To a mixture of 180 grams of commercially available molybdenum powder, 4.1 micron Fisher number (maximum metallic impurities .Ol5%) with 5.76 parts of methyl cellulose (Methocel 60 H.G.) were added 20 parts of 10 percent glycerin-water solution. Using a sigma blade mixer, the ingredients were thorougly mixed to a claylike mass. This mixture was then milled on a rubber mill to a sheet of approximately 50 mils thickness. The sheet was dried overnight at 66 C.

After drying, the green sheet was cut into a number of pieces suitable for testing and then sintered for about 24 hours at 1100 C.-in dry hydrogen, followed by 1 hour at 1800 C. in a vacuum furnace at a pressure of approximately l 10- mm. Hg. When tested on an Instron tester, using a rate of loading of 0.02 inch/min, gauge width of 0.25 inch and gauge length 1 inch, the following results were obtained with the sintered sheet. The tests were carried out at about 25 C.

Specimen 3 was held at 1,650 C. at approximately 1X10 mm.

Hg for 1 hour after its initial sintering at 1,800 0.

Specimens made in this way could be bent at room temperature to an angle of approximately 90 over a 1.4T radius at the rate of 0.02 inch/min. without failure. (T -thickness of sheet being tested.)

Similar specimens were made, where the green sheets were densified by pressing prior to firing. The green sheets were pressed at 30,000 p.s.i. at room temperature, whereupon a thickness reduction of approximately percent took place. The pressed sheets were then held for 40 hours at 500 C. in dry hydrogen and then fired for 1 hour at 1800 C. at approximately 1x10- mm. Hg. Upon testing as set forth above the following values were obtained.

TABLE II Specimen Ultimate Tensile Elongation,

Strength, p.s.i. Percent Specimens made in this way could likewise be bent without breaking at room temperature.

All of the specimens prepared as set forth above had carbon content of about 0.37 percent by weight on analysis.

When the above process was repeated, using a fugitive binder, comparable samples were obtained, except that they contained no analytically demonstrable carbon. Thus, for example, the procedure was repeated, using polymethyl methacrylate as the binder and nitroethane as the solvent, and an identical sintering cycle was followed. The specimens produced in each case were brittle on bending and could not be bent without fracturing. They had no ductility at all at room temperature.

Example 2 Green sheet material of identical composition to that set forth in Example 1 was pre-sintered in hydrogen for 12 hours at 1100 C. The resulting molybdenum sheets were then sintered as follows:

C. in vacuum microns). A number of these specimens were further sintered for 1 hour at 2000 C. at 4 l0- mm. Hg. Strong, dense sheets of tungsten were obtained.

The sheets made by the process disclosed herein are significantly stronger after heating than the similarly exposed commercial material. This is amply demonstrated by the test results set forth in the following table. These were found after heating specimens of commercial tungsten sheet and comparable specimens made according to this example. All specimens were about inch wide and mils in thickness and were first exposed to 2000 C. for 1 hour in vacuum prior to testing, then cooled to room temperature and then heated to test temperature as shown. Specimens designated A were commercial sheets; those designated B were made according to this example.

TABLE IV Specimen Test Temperature, Modulus of Rupture,

1 Average of 2.

(Specimens were subjected to bending using a 1 inch span and a load rate of .02 inch per rninulte. The series of temperatures used bracketed the brittle-ductile transition temperature of the specimens.) The A series of specimens contained less than 0.01 percent of carbon. The carbon content of the B series of specimens was 0.04 percent.

Example 4 To a mixture of 4000 grams of tungsten powder (.85 micron Fisher sub-sieve size, ASTM B33058T; comcercially available; maximum metallic impurities .0l42% with 80 grams of methyl cellulose (Methocel 60 H.G.) were added 400 cc. of 10% glycerin-water solution. Using a sigma blade mixer, the ingredients were thoroughly mixed to a clay-like mass. A portion of this mix was milled into a sheet approximately mils thick. This green sheet was dried overnight at 68 C. in air and then pressed at 30,000 p.s.i.

After sintering by heating to 2000 C. over a period of 2 hours in vacuo (5 1O- mm. Hg) and holding at that It can be seen that exceptional densities are achieved with this relatively coarse powder. Photomicrographs of metallographically mounted and polished specimens showed the presence of small amounts of carbon.

Example 3 To a mixture of 1500 grams of commercially available tungsten (1.45 micron Fisher number, maximum metallic impurities 0.03l4%) with 30 grams of methyl cellulose (Methocel H6.) were added 190 cc. of 5 percent glycerin-water solution. Using a sigma blade mixer, the ingredients were thoroughly mixed to a clay-like mass. This mixture was then milled on a rubber mill to form a sheet of approximately 50 mils thickness. The sheet was dried overnight at 66 C. and then Was pressed between flat plates atabout 37,000 p.s.i. Specimens were cut from the sheet and sintered for 3 hours at 1300 temperature for 1 hour, dense, strong sheets of tungsten were produced. Compacts were prepared from the same powder using 30,000, 50,000 and 100,000 p.s.i. and no binder. These compacts were sintered in an identical manner to that of the green sheets. The above specimens as well as specimens of commercial tungsten sheet were heated over a period of 2 hours to 2000 C. under reduced pressure (less than 5X l0- mm. Hg) and held at that temperature and pressure for 1 hour. A grain size comparison revealed that the material produced according to the process of the invention has much finer grain size than the commercial sheet or the compacts made without binder. Specimens of the sheets of the invention, the compacts made from powder and commercial tungsten sheet were again exposed to 2000 C. under the conditions described for 1 hour to determine the effect on grain size and density.

The metallic sheets of the invention remained of exceedingly fine grain size, whereas the commercial material and the powder compacts made without binder changes to markedly coarse grain size.

The densities of sintered specimens produced with and without binder were as follows:

8 if it is to be extruded or compression-molded. The integrity of the shape or extrusion is aided, however, if the green mass is held under pressure of about 1X10 mm. Hg or less for a period sufficient to remove occluded gases prior to extruding or molding. After forming into the desired configuration, the green structure is carefully 1 Very low-carbon content of the tungsten powder is of the order of p.p.m.

Example 5 A portion of the clay-like mass of Example 4 was placed in an extrusion press. A splitter die was used and the material was extruded into a tube of approximately 2.4 inch OD. and 0.100 inch wall.

This green tube was dried in the extruded position and cut into two lengths about 10 inches long and sintered (using graphite mandrels of 1.655 diameter) at 2000 C. for one hour. It is necessary when sintering to heat slowly from 180 C. to 350 C. to permit decomposition products from the binder and plasticizer to evolve without rupture of the green mass. Seamless tungsten tubes having high density and strength similar to that described above are obtained.

Example 6 A clay-like mass of tungsten, methyl cellulose, glycerin and water produced as described in Example 4 was formed into a sheet on a rubber mill, using a roll-roll speed ratio of 1.4 to 1. This sheet containing virtually all of the plasticizer, binder and solvent is extremely flexible and was cut into a number of smaller square sheets measuring several inches along each side. These were corrugated (7 nodes per inch) on a steam heated roll corrugator. The corrugated sheets were allowed to dry and then the nodes were moistened using a 1 percent methyl cellulose-water solution, and the sheets were stacked on each other, corrugations being joined node to node, the directions of the corrugations of alternate sheets being at right angles. The green, temporarily adhered intermediate structure was fired by the sintering process described in Example 3. An integral article consisting of corrugated tungsten sheets sinter-welded together with no evident inhomogeneity at the welds on microscopic examination of metallurgical specimens was produced.

Similar shapes are made using machined dies to press the corrugations. Die pressing has been found to be an excellent means of producing corrugations of high dimensional tolerance.

Example 7 A mixture of molybdenum powder, methyl cellulose and water made by the process of Example 1 was milled on a rubber mill with a roll speed ratio of 1.4 to 1 into a flexible green sheet of 0.040 inch thickness. This sheet was corrugated and fabricated into shapes of various configurations by the technique described in Example 6. Green material made in the identical manner was extruded into tubes and injection or compression molded into various dies.

It is not necessary to mill the mass of green material sintered first at 1100 C. in dry hydrogen or in vacuo, then at 1800 C. in vacuo, as set forth in Example 1. Strong molybdenum articles are formed. Where these are composed of several pieces joined together, they are found to be sinter-welded at the joints, and no discontinuity or inhomogeneity of the metal at these joints can be detected.

' Example 8 Commercial molybdenum powder with the characteristics shown in Table VI was used to prepare sintered rods from which tensile specimens were machined. Green rods of 0.30" diameter were prepared by extruding an aqueous, plastic mass of methyl cellulose binder and metal powder prepared using the method of Example 1. These rods were dried, iso-pressed at 30,000 psi, and subjected to thermal treatments designed to break down the binder to carbon and to control the amount of residual carbon present in the body prior to sintering.

Three different conditions were employed in which rods were heated to 1200 C. in (a) Wet hydrogen, (b) dry hydrogen and (c) vacuum. Subsequent vacuum sintering (5 10 torr) of these three groups of samples at 1800 C. for two hours gave rise to specimen groups A, B and C in Table VII, corresponding to carbon contents of 86 p.p.m., 3100 p.p.m. and 5600 p.p.m., respectively.

Tensile test specimens with 1.0" gauge length and 0.10" gauge diameter were ground from sintered rods and electropolished prior to tensile testing. Room temperature tensile testing was carried out at a loading rate of 0.05 cm./min. on an Instron testing ,machine equipped with extensometers.

TABLE VI.MOLYBDENUM POWDER CHARACTE RISTICS Chemical Analysis Element Impurity Element Impurity (p-p- (p-p- 1 Powder commercially available from the General Electric Corp.

Particle Data.Average particle diameter (Fisher subsieve slzer number), 3.8 microns; Porosity 0.635. Apparent density (ASTM B32958), 30.4 g./in.

TABLEVII.ROOM-TEMPERATURE TENSILE DATA FOR UNWORKED SINTERED MOLYBDENUM Elong. of A Analysis (p.p.m.) Density T.D. Hardness Grain (p.s.i.) (p.s.i.) (percent) (percent) (g./cm. (Percent) DPN size 02 N2 (mm.)

1 Ultimate Tensile Strength=Loadloriginal area at fracture. 2 Upper Yield Point.

Example 9 Rod suitable for preparation of tensile specimens was produced as follows:

The materials and quantities employed are: 5000 g. of commercial tungsten powder, .8 micron Fisher number; 125 g. of Methocel (60 H.G., 400 c.p.s.); and 400 ml. of ten volume percent aqueous glycerin solution. The tungstem and Methocel were dry mixed in a sigma blade mixer for one hour. The water-glycerin solution was then slowly added, and a very stitf clay was produced within minutes. The batch was allowed to mix for five minutes under vacuum to become de-aired.

Atmosphere: Hydrogen bubbled through water at room temperature.

Heating rate: 25 C. to 1250 C. in eight hours.

Time at temperature: Two hours, 1250 C.

sintering- Atmosph'ere: Vacuum, final, 1 10- torr.

Heating rate: 25 to 1250 C. in one-half hour;

1250 C. to 1950 C. in two hours.

Time at temperature: Two hours, 1950 C to Specimens were prepared from sintered rod by grinding to one inch gauge length and 10 mil diameter. The test specimen was then tapered approximately .001 from end to center using abrasive papers. The specimens were electropolished using two percent aqueous NaOH solution. Specimens were pulled using an Instron machine at a loading rate of 0.05 cm./min. All elevated temperature testing was performed in vacuum. Deformation was measured using an extensiometer attached to the grips.

Carbon was determined by volumetric methods.

Density was determined using a mercury penetrometer.

Tables VIII and IX list the data obtained from the density, tensile and hardness tests.

TABLE VIII.VAOUUM SIN'IERED TUNGSTEN Sintered Carbon Diamond Test 2% Ofiset Spec. No. Density 1 Content Pyramid Temp., U.T.S. Yield Stress Elong.

(gnL/Cc.) (p.p.m.) Hardness, C. (p.s.i.) (p.s.i.) (percent) 1 Green density 9.4 gmJcc. 2 98.3% Theoretical.

TABLE IX.WET HYDRO GEN PREFIRED-VACUUM SINTERED TUNGSTEN 1 Green density 9.4 gmJcc. 2 93.6% theoretical.

The ale-aired material was charged into an extruder and again evacuated for two minutes. The vacuum was maintained until the rod began to extrude. The green rod was 0.30" in diameter, and was dried at C. for 16 hours and then iso-pressed at 30,000 p.s.i.

All specimens used were made from the same lot of tungsten powder and the same batch of green material.

The following sintering cycles were employed.

(1) Vacuum series (specimens in Table VIII):

Atmosphere: Vacuum, final, 1 10- torr. Heating rate: 25 C. to 1950 C. in three hours. Time at temperature: Two hours, 1950 C. to 2000 (2) Wet hydrogen series (specimens in Table IX):

It can be seen from the tables that the density of the vacuum sintered series (Table VIII) is higher than the density of the samples prefired in wet hydrogen (Table IX). The vacuum sintered material contains 2800 p.p.-m. carbon that originates from the thermal degradation of the binder, whereas the specimens prefired in wet hydrogen contain less than ppm. carbon.

It is also apparent that the strength of the tungsten containing carbon is much greater than that of the tungsten from which the carbon has been removed. The carbon is present predominantly as tungsten carbide. As with molybdenum, alloys of tungsten with other metals can be produced as described herein, containing carbon as metal carbides and having similarly improved properties.

A review of the literature concerning the strength of 1 1 wrought tungsten in the recrystallized condition will reveal that the strength values of Table VIII are high and in most instances exceed literature values. This is regarded as particularly significant considering that the specimens of Table VIII have never been mechanically worked.

Example Specimen rods were prepared from commercial tungsten powder characterized in Table X as such or in admixture with binders and other materials. The technique used was to fill a cylindrical rubber mold to form a rod approximately /2 inch in diameter and 4 inches long. The powder was tamped with a /2 inch diameter stainless steel rod to increase the density. The mold was then evacuated with a mechanical pump and sealed. The specimens from untreated powder are marked I.

Water-treated tungsten powder specimens (II) were prepared by placing tungsten powder in a sigma blade mixer and adding 7.2 cc. of distilled water per 100 grams of powder while mixing. The material was mixed for one hour, or until the powder was completely wet. It was then placed on a glass pan and dried at 150 F. for 12 hours. No grinding operation was necessary after the drying since the material was still in the form of a powder. In a like manner specimens of tungsten powder treated with water and glycerin (III) were prepared, using 8 cc. of 10 percent glycerin solution in distilled water per 100 grams of tungsten powder.

Specimens containing tungsten and graphite powder (IV) were prepared in a twin-shell blender, using 0.6 gram of microcrystal graphite (grade No. 1651, Southwestern Graphite Company) per 100 grams of tungsten powder. The mixture was blended for two hours.

Specimens containing tungsten powder, glycerin, water and Methocel (V) were prepared by first mixing dry 2.5 grams of methyl cellulose (60 H.G., 4000 c.p.s.) per 100 grams of tungsten powder in a twin-shell blender. The tungsten-methyl cellulose mixture was then placed in a sigma blade mixer, and 8 cc. of 10 percent glycerin in distilled water solution were added for each 100 grams of tungsten powder. After mixing an hour the material was a clay-like mass. The mixer was then evacuated to remove air. The clay was then placed in a small extruder, and a rod of diameter approximately 0.3 inch was extruded. The rod was placed on a V-rack under a damp cloth, dried for 12 hours at about 40 F., heated to room temperature, then dried at 150 F. for two hours. The rod was then cooled to room temperature and inserted in a cylindrical rubber mold, which was then evacuated and sealed.

All specimens were iso-pressed in the molds at 28,000

TABLE X.TUNGSTEN POWDER ANALYSIS 1 Impurity Content Element Impurity Element Impurity (p.p.m.) (p.p.m.)

1 Powder commercially available from the General Electric Co.

Particle Data-Average particle diameter (by Fisher sub-sieve size number), 0.80 micron; Porosity 0.765. Apparent density (ASTM B-32958), 29.0 gm. per cu. in.

The specimens thus prepared were fired as follows:

Hydrogen prefiring, both wet and dry, was done in an alumina tube resistance furnace. The wet hydrogen was bottle hydrogen bubbled through tap water at room temperature, and the dry hydrogen was bottle hydrogen that 12 was passed through a liquid nitrogen trap. The gas flow was approximately 0.025 c.f.m. in'the two inch diameter furnace tube. The heating cycle was controlled automatically and recorded on a chart recorder. The specimens were heated to 1200 C. in two hours at a nearly constant heating rate, held at 1200 C. for two hours, then allowed to cool to room temperature in the hydrogen atmosphere.

The vacuum prefiring and final sintering (all done in vacuum) were performed in a 30 kw. induction furnace, using a molybdenum susceptor and shields. A mechanically backed diffusion pump was used for evacuation. The prefire cycle was the same as for the hydrogen prefire, except that it was controlled manually. The final sinter cycle consisted of heating to 2000 C. in three hours, and holding at 2000 C. for two hours. Power was then shut off and the specimens cooled in vacuum. All compositions were fired separately except I and II, which were prefired and final-sintered together.

Samples were taken from the green and sintered specimens for analysis for carbon and oxygen. The green and analytical samples were cut from the specimens with a hacksaw, and the sintered samples were chipped from the specimens with a cold chisel. All of the sampling was done dry to avoid possibly introducing impurities that would afl ect sintering or analysis.

Density measurements were made on green and sintered samples. To seal the green samples from absorbing liquid during the measurements, they were coated (by immersion) with two coats of collodion. A thin wire was attached and the samples were weighed in air, then in glycerin. The density measurements for the sintered samples were made with a mercury penetrometer.

The green and sintered densities of the tungsten specimens and oxygen and carbon after sintering are given in Table XI.

TABLE XL-GREEN AND SINTERED DENSITIES OF TUNGSTEN ROD Green sintered sintered Sintered Sample and prefire density density carbon oxygen atmosphere (percent (percent content, content,

theoretical) theoretical) p.p.m. p.p.m.

I. Wet hydrogen 48.0 87.9 5 15, 28, 40 1. Dry hydrogen 48.0 92.0 4 27 I. Vacuun1 48.0 88. 2 7 34 II. Wet hydrogen 49. 3 86. 7 2 30 II. Dry hydrogen 49. 3 88.4 4 38 II. Vacuum 49. 3 79. I 9 40 III. Wet hydrogen 51.3 89.1 4 6 III. Dry hydrogen 51. 3 93. 6 172 18 III. Vacuum 51.3 84.4 14 38 IV. Wet hydrogen 46. 6 87.1 12 77 IV. Dry hydrogen..- 46. 6 94. 6 3, 2 N.D. IV. Vacuum 46.6 94.4 1, 650 1 N .D V. Wet hydrogen 49. 5 92. 8 11 8 V. Dry hydrogen 49. 5 96. 2 1, 260 V. Vacuum 49. 5 96. 7 2, 230 2 N.D

1 P.p.m.=Parts per million by weight. 2 N.D.=not detectable, detection limit is 3 p.p.m.

From this experiment, it can be seen that the sintered specimens with high carbon content have a very low oxygen content. In some cases the oxygen has been reduced to non-detectable limits.

It can also be observed that the sintered density of the specimens containing carbon is considerably higher than ;he density of the specimens that are relatively carbon ree.

It will be noted that carbon added as a very fine (.6 micron) graphite has a beneficial effect on sintered density. However, addition of particulate carbon is not particularly advantageous for several reasons:

(1) Although improved, the density is not as high as obtained from the use of a degradable binder.

(2) Free graphite or carbon remains in the structure. Photomicrographs of metallographic mounts of the vacuum-sintered tungsten and graphite specimen at 500 show carbonaceous material (probably free graphite) that colors the abrading surface black.

(3) The grain size is not as fine as observed in the vacuum-sintered tungsten-methyl cellulose specimen.

The vendor analysis of the oxygen content of the pow- 13; der was 1620 p.p.m., but analysis of the as-received tungsten showed the oxygen content to be 3200 p.p.m. Tungsten specimens prepared as described in this example were 14 bon to obtain desired results. All specimens were sintered at approximately 2000 C. for two hours at a pressure of not more than 1X mm. Hg.

1 Specimens prepared and tested using the procedure of Example 9. 2 Mercury penetrometer data. 3 Based on analyses of sintered rods trom identical powder in Example 10.

analyzed in the green state after drying, but before sintering, with the following results.

It can be seen that the use of aqueous systems has the effect of highly oxidizing the tungsten powder used. It is therefore significant to realize that vacuum sintering of these systems in the absence of carbon does not reduce the oxygen content below about p.p.m., but that the oxygen content is reduced to non-detectable (less than 3 p.p.m.) amounts if carbon is present. Thus the present invention has the further advantage of allowing the use of highly oxidized powders.

(4) Particulate carbon or graphite does not act as a binder and provide the configurational freedom of the organic binder materials.

Example 11 Another experiment was conducted to determine the effect of different carbon levels on strength and oxygen content of tungsten. Specimens of tungsten rod were prepared using methylcellulose and glycerin-Water (designated A) as described in Example 10. Specimens of pure tungsten rod (designated B) were prepared by filling the molds with pure tungsten powder. The specimens were iso-pressed at 28,000 p.s.i. The prefire cycles used and the resulting properties are shown in Table XII. Normal wet and dry hydrogen prefire atmospheres were used but the specimens were not held at the prefire temperature for the normal 2 hours. Upon reaching temperature, power was shut 011 and the specimens allowed to cool. In this manner it was possible to reduce carbon levels appreciably using wet hydrogen, yet maintain sufficient car- The beneficial effect of the carbon on density is again demonstrated. The beneficial effect on the tensile strength of carbon from binder degradation can also be seen.

What is claimed is:

1. An unworked sintered article consisting essentially of molybdenum or molybdenum alloy, the said molybdenum containing from 0.04 to 0.80 percent of carbon homogeneously distributed therein predominantly as metal carbides of the Group VI metals and being ductile in bending and tensile testing by loading on a one inch gauge length at 0.02 inch/ minute at room temperature.

2. An unworked molybdenum or molybdenum alloy structure according to claim 1, having an average grain size diameter ranging from extremely fine up to about 0.025 millimeter and retaining substantially the same grain size after heating to 1800 C. for one hour in vacuum.

3. An unworked molybdenum or molybdenum alloy article according to claim 1, exhibiting ultimate tensile strength over 65,000 p.s.i. at room temperature.

4. An unworked sintered article consisting essentially of tungsten or tungsten alloy, having grain size ranging from extremely fine up to about 0.01 millimeter and density greater than about 97 percent of theoretical containing 0.04 to 0.80 percent of carbon homogeneously distributed therein predominantly as metal carbides of the group VI metals.

5. An unworked tungsten or tungsten alloy article according to claim 4, exhibiting ultimate tensile strength over 45,000 p.s.i. at room temperature.

6. An unworked sintered article consisting essentially of tungsten or tungsten alloy, or molybdenum or molybdenum alloy, having grain size in the case of tungsten or alloy thereof ranging from extremely fine up to about 0.01 millimeter and density greater than about 97 percent of theoretical, containing 0.04 to 0.80 percent of carbon homogeneously distributed therein predominantly as metal carbides of the group VI metals.

(References on following page) References Cited UNITED STATES PATENTS Greeger 75201 X Timmons 75176 Redden 75176 Grant 29182.5 X

Keith 29-182.5 Semchyschen 75176 Dickinson 75--176 Chang 75176 Del Grosso 75-203 907,789 10/1962 Great Britain.

OTHER REFERENCES 5 The Metal Molybdenum," ASM publication, 1958,

EflFect of Carbide Dispersion in Molybdenum Alloys, Chang (II), Trans. AIME, April 1960, yo]. 218, pp. 254 256.

CARL D. QUARFORTH, Primary Examiner.

A. J. STEINER, Assistant Examiner.