US3551991A - Infiltrated cemented carbides - Google Patents

Abstract

CEMENTED CARBIDE COMPOSITIONS ARE PREPARED BY THE PRODUCTION OF A SINTERED CEMENTED CARBIDE COMPACT BY CONVENTIONAL LIQUID-PHASE SINTERING TECHNIQUES, FOLLOWED BY CHEMICAL REMOVAL OF THE METALLIC BINDER AND REPLACEMENT BY INFILTRATION WITH A SECOND METALLIC BINDER DIFFERENT THAN THE ORIGINAL BINDER. THE RESULTING COMPOSITIONS ARE FULLY DENSE, HAVE A FINE-GRAINED, HOMOGENEOUS STRUCTURE AND POSSESS THE HIGH HARDNESS NORMALLY ASSOCIATED WITH CEMENTED CARBIDE ALLOYS.

Description

United States Patent 3,551,991 INFILTRATED CEMENTED CARBIDES William A. Reich, Huntington Woods, and Thomas E. Hale, Warren, Mich., assignors to General Electric Company, a corporation of New York No Drawing. Filed Apr. 16, 1969, Ser. No. 816,778 Int. Cl. B22f 3/24 U.S. Cl. 29420.5 9 Claims ABSTRACT OF THE DISCLOSURE Cemented carbide compositions are prepared by the production of a sintered cemented carbide compact by conventional liquid-phase sintering techniques, followed by chemical removal of the metallic binder and replacement by infiltration with a second metallic binder different than the original binder. The resulting compositions are fully dense, have a fine-grained, homogeneous structure and possess the high hardness normally associated with cemented carbide alloys.

This invention relates to a process for the production of cemented carbide compositions having an ususual combination of desirable properties and to the cemented carbide compositions so produced.

Cemented carbides are well known for their unique combination of hardness, strength and abrasion resistance and are, accordingly, extensively used for such industrial applications as cutting tools, drawing dies and wear parts. They are produced by powder metallurgy techniques involving the liquid-phase sintering of one or more refractory carbides of Groups IV, V and VI of the Periodic Table with one or more of the iron-group metals. The iron-group metals exist as a matrix or binder in the sintered alloy and act to bond or cement the refractory carbides.

The matrix alloy is almost always an iron-group (Fe, Co, Ni) metal or alloy because the latter possesses a number of critical properties necessary to the sintering process and to the metallurgical structure upon which the unique properties of the sintered alloy depend. To accomplish the desired liquid-phase sintering, the binder metal must have the ability to produce a reasonably low melting point with the carbide through eutectic formation. It must also possess the ability to dissolve a fairly large amount of the carbide while in the liquid state and it must adequately wet the carbide such that densification is accomplished completely and within a reasonable period of time. In addition, the binder should not enter into an irreversible reaction with the carbide to any great extent. Finally, the binder should contribute to the high strength of the cemented carbide alloy, preferably at elevated temperatures as Well as at room temperatures. As a result of these complex demands, there are only a limited number of binder metals commercially used. Cobalt is by far the most extensively used binder for tungsten carbide, and nickel is the most extensively used for titanium carbides.

These alloy systems display the unique properties normally desired in cemented carbides. However, they also have certain limitations. One of the principal disadvantages of cobalt as a binder for tungsten carbide alloys is its relatively poor corrosion resistance, particularly to acids. Another disadvantage is the relatively low resistance to deformation at elevated temperatures of tungsten carbide alloys in which cobalt is the binder.

Many other binder metals or alloys have been suggested, particularly as a substitute for cobalt in WC-Co alloy systems. However, many are unsuitable for reasons set forth above. In other cases, such metals or alloys may be used as binders but only in volume percentages greater than about 15 or 20%. For example, US. Pat. 3,215,510 discloses the use of nickel-chromium alloys as a substitute for cobalt in WC alloys, with resulting improved corrosion resistance without sacrifice of strength or hardness. Satisfactory densification and other properties in the sintered compact are obtained, however, only with binder contents greater than about 15 volume percent 10 weight percent).

It is a principal object of this invention to provide cemented carbide compositions containing a matrix phase of a metal which would not, by known liquid-phase sintering techniques, produce a fully dense, fine-grained, homogeneous structure. It is an additional object of this invention to provide cemented carbide compositions possessing a combination of properties heretofore unattainable or attainable only with great difiiculty. It is an additional object of this invention to provide a process for producing such compositions.

We have found that the foregoing objects may be achieved by preparing a cemented carbide alloy of a refractory metal carbide in combination with an iron-group metal as a matrix by well-known powder metallurgy techniques. This produces a fully dense cemented carbide structure of acceptable properties via liquid-phase sintering. The cobalt or other iron-group metal binder is then removed from the cemented carbide alloy by elec trochemical or chemical means, as for example by leaching with an acid, to produce a carbide skeletal structure. This skeletal structure is then infiltrated by a second binder metal different from the original binder and possessing the properties desired in the final composition. In this manner, the inherent advantages of the cobalt or nickel matrix are utilized to prepare a cemented carbide of the desired density, structural uniformity and homogeneity, after which the original binder is then removed and replaced with the selected matrix containing the properties desirable in the final cemented carbide composition. Stated otherwise, the cobalt or nickel binder is used as a processing aid because of its uniquely suitable prop erties for this purpose. The products of this invention are fully dense-greater than 99% and usually greater than 99.5% by volume of theoretical density, have a hardness greater than about and usually greater than Rockwell A, are strong and are characterized by a fine grained, homogeneous metallurgical structure.

Using this process, it is possible to produce cemented carbide alloys which contain binder metals or alloys which do not themselves satisfy the stringent metallurgical requirements for liquid-phase sintering. The only metallurgical requirements that the binder phase must possess for this method are (1) the ability to wet the carbide skeleton well enough to infiltrate the skeleton, and (2) freedom from detrimental irreversible reactions with the carbide. There exist many metals and alloys which satisfy these reduced requirements but which do not possess the additional properties necessary for liquid-phase sintering. For example, by this process it is possible to make WC, TiC, or WC-TiC-TaC-based composites having fine (1 to 3 micron) uniformly dispersed carbide grains at a level of greater than 80 volume percent with a matrix phase of elements such as copper, silver or gold, or of alloys such as nickel-chromium, nickel-aluminum, cobalt-aluminum, copper-silicon, alloys of nickel or cobalt and refractory metals such as niobium, tantalum, chromium, molybdenum or tungsten.

The invention also makes possible the preparation of fully dense cemented tungsten carbide compositions of high corrosion resistance containing less than 15% by volume of a nickel-chromium binder alloy. As set forth above, this has not been previously possible with a volume of binder less than 15%. The practice of this invention also makes possible the preparation of cemented carbide alloys containing nickel-aluminum alloy binders with from 1 to by weight aluminum. Such binders are well known for their high temperature strength and would, accordingly, enhance the high temperature deformation resistance of a cemented carbide. Again, they have been unsuitable because of their inability to participate satisfactorily in the liquid-phase sintering process to obtain the required densification and fine-grained, homogeneous structure. The high temperature strength of cobalt is also enhanced with from 1 to 10% aluminum. Moreover. it becomes possible to prepare carbide alloys with aluminum alloy binders of less than 20 volume percent. This has been extremely clifficult because a substantial portion of the aluminum in the binder alloy irreversibly oxidizes during normal cold-pressing and liquid-phase sintering processing. The result has been the formation of numerous oxide inclusions in the cemented carbide, which is detrimental to the properties of the resulting cemented carbide composition. The present process avoids substantially all aluminum oxide formation because the aluminum need not be present as a powder at any stage of the process.

Additional binder alloy systems made possible by the practice of this invention are nickelor cobalt-based binder alloys having higher aluminum contents, such as the nickel-aluminum compounds Ni Al, NiAl and the cobalt compound CoAl. These nickel and cobalt intermetallic compounds are attractive as binders for both tungsten carbide and mixed tungsten carbide alloys as a means to obtain exceptionally high deformation resistance temperatures--up to 1350-1600 C. It is not possible to make fully dense cemented carbides having less than 20 volume percent of such binders by liquid-phase sintering since the solubility of WC in these compounds is very I low, preventing the redistribution of the carbide phase which is so necessary in the liquid-phase sintering process.

The cemented carbide composition should contain sufficient volume percent of refractory carbide metal to leave an intact skeleton after removal of the binder. For example, in the case of WC-Co and WC-TiC-TaC-Co compositions, the carbide skeleton will usually break into several pieces or disintegrate into powder during the leaching process when the carbide content is less than about 80 volume percent. It is possible to prepare carbide skeletons by other means, such as hot-pressing, which would not disintegrate at a volume percent of less than 80%. Hot-pressing can also be employed to obtain carbide skeletons which have greater than 80% volume density. However, hot-pressing has a number of disadvantages. It is a slow and expensive process, since only a few parts can be made at a time and since the graphite molds which must be used do not last long under the conditions required to obtain the desired density. In addition, it is difiicult to obtain, by hot-pressing, uniform density and grain size throughout the cross section of the skeleton, and the surface is usually impregnated or contaminated with the mold material.

Using the process of this invention, large quantities of carbide skeletons can be prepared simultaneously using a conventional sintering furnace and a leaching tank of sutficient size, The resulting skeleton quality is excellent in terms of grain size, uniformity of density, and freedom from surface contamination. The process thus offers cost and, even more important, quality advantages over hot-pressing for preparation of the skeletal carbide body.

For some applications, it may only be necessary to remove the cobalt binder from the surface regions of the initial cemented carbide body. In this case a different or second binder metal can be infiltrated into the carbide to any desired depth, in order to provide the desired properties in this critical surface region, while the cobalt binder remains in the bulk of the carbide body. By varying the time and temperature of infiltration, varying degrees of alloying can be obtained between the infiltrated metal and the bulk cobalt binder. When the binder is to be replaced only in the surface region, the infiltrant is restricted to those metals and alloys having a melting point lower than that of the WC-Co or WC-TiC-TaC-Co eutectic. This is necessary to prevent the bulk binder from melting and migrating into the surface carbide skeleton before the desired infiltrant has had the opportunity to do so. Infiltrant alloys which dissolve the carbide should be presaturated with the carbide in order to prevent dissolution of the skeleton. Thus nickel-chromium alloys should normally be saturated with tungsten carbide when infiltrating a tungsten carbide skeleton.

Conventional liquid-phase sintering comprises coldpressing an intimate powder mixture of refractory metal carbide and binder metal to the desired shape. The pressed part is then heated, normally in an inert or reducing atmosphere, to an elevated temperature, normally from l300-l500 C., to sinter the alloy. At the sintering temperature, a liquid phase of the binder metal and a small amount of the refractory carbide forms, permitting complete and rapid densification.

After the cemented carbide has been cold-pressed and sintered, the binder metal is removed. We have found this may be conveniently and economically accomplished by leaching with an acid. To accelerate the leaching process, the acid should be heated, most conveniently to boiling temperature. Acids which are particularly useful are hydrochloric and sulfuric, preferably at dilute constantboiling concentrations. Leaching will normally occur over a period of days or weeks for complete removal of binder, depending of course on the size of the product, or lesser time if only the surface is to be leached.

After leaching has proceeded to the desired extent, the residual acid and water are removed, as by firing in a hydrogen atmosphere. Frequently, the carbide skeleton is oxidized to some extent during leaching, the net result of which is a loss of carbon from the skeleton. When such a carbon loss would be detrimental, the carbon can be replaced by firing the skeleton in a carburizing atmosphere and/or by adding the desired carbon to the infiltrant.

Infiltration is carried out in a reducing or slightly carburizing atmosphere, depending on the specific composition of the carbide skeleton structure. The infiltrating binder, in such forms as a powder, solid body, or foil, is raised to slightly above its melting point while in contact with the skeleton and then infiltrates by capillary action.

The invention will be more clearly understood from the following examples. In these examples (and elsewhere in the specification), proportions of carbide and binder phase are, unless otherwise indicated, by volume to give more truly comparative values because of wide differences in binder densities, whereas constituents of the binder itself are in percent by weight of the binder alloy, in accordance with conventional practice for metallic alloys.

EXAMPLE 1 Rectangular-shaped pieces of cemented carbide having dimensions .750" x .375" x .200" and composition volume percent WC1O volume percent Co (6 weight percent Co) were prepared by conventional cold-pressing and liquid-phase sintering techniques. The pieces were then leached for seven days in boiling 20% hydrochloric acid to remove the cobalt. At the end of this period a chemical analysis showed the residual cobalt content to be 0.16 weight percent. The resulting tungsten carbide skeletons were then fired one hour at 1100 C. in a carburizing hydrogen-base atmosphere to remove residual acid and to help restore carbon that was lost due to oxidation during leaching. The tungsten carbide skeletons were then infiltrated with nickel-chromium alloys containing additionally 20% WC and 2% carbon by weight. The WC was added to presaturate the infiltrant with WC so as to prevent erosion of the WC skeleton during infiltration. The carbon was added to restore the carbon content of the WC skeleton and thus prevent the formation of eta phase type ternary carbides. The amount of infiltrant used was by weight in excess of the amount needed to fill the voids in the carbide skeleton.

ton and heating in a hydrogen or vacuum atmosphere to a temperature that was about 25 C. over the melting point of the infiltrant. The temperature used ranged from 1450 C. for the 5 to aluminum-containing alloys to 1700 C. for the 30% aluminum alloy.

Infiltration was accomplished by placing a premelted 5 The following Table II illustrates the physical properbutton of the infiltrant alloy over the carbide skeleton, ties observed for this alloy series (Compositions 3-6) raising to a temperature of 1425 C. in a hydrogen atplus those for the original WC-lO volume percent Co mosphere and holding minutes at 1425 C. The resultcomposition (Composition 1) and for a bar infiltrated ing compositions were found to have excellent physical 10 with pure nickel (Composition 2). Relative to the strength properties and superior corrosion resistance as shown in of the original WC-Co material the strength of the Ni-Al the following Table I. Included in this table are corrematrix compositions was seen to rise from a value somesponding properties of a WG-10 volume percent Co comwhat lower than the original at 0% Al to a value submercial alloy (Composition 1) from which the WC-Ni, stantially higher at 2.5% A1 and then steadily decline Cr alloys (Compositions 2, 3 and 4) were prepared: 15 with increasing aluminum content. The hardness was rela- TABLE I Transverse rupture Corrosion Density, strength, Hardsuscepti- Composition gm./cc. p.s.i. ness, RA bility '1. WC10 volume percent 00 14. 95 309, 000 92. 0 258 2. wo-lo volume percent (Ni-5 Cr) 14. 98 255, 000 92.0 6.3 a. WG-10 volume percent (Ni-2.5 Cr) 14. 98 318,000 91. 9 8.6 4. WG-10 volume percent (Ni- Cr) 14. 82 250,000 92. 5 14. 7

N aCl-acetic acid solutionweight loss in milligrams per square decirneter per day.

As can be seen, the density, hardness and strength of the WC-Ni-Cr alloys are nearly the same as the WC-Co tively low at 0% Al and then rose steadily with increasing aluminum content.

*Extrapolated values.

7 alloy, while acid corrosion resistance is substantially in- EXAMPLE 2 WC skeletons having 10 volume percent porosity were prepared by acid-leaching the cobalt from WC-10 volume percent Co cemented carbide composites and were subsequently hydrogen-fired as in Example 1 above. These were then infiltrated with a series of nickel-aluminum alloys containing from 2.5 to 30 weight percent aluminum. Suificient WC and carbon were added to each infiltrant to prevent erosion and to prevent the formation of eta phase type ternary carbides. This amounted to 5% WC when the aluminum content was from 2.5 to 10% and no WC when the aluminum content was greater than 10%. Two percent carbon was added to all compositions.

Infiltration was accomplished by placing premelted buttons of the desired infiltrant alloy over the WC skele- Evaluation of the alloys listed in Table II indicated that the higher aluminum-containing alloys have better resistance to high temperature deformation than comparable WC-Co alloys. For example, a bar of the WC-10 volume percent (Ni-18% Al), Composition 5 in Table II, was found to undergo essentially no short-time plastic deformation when subjected to 6000 p.s.i. pressure at 1300 C., whereas the comparable WC10 volume percent Co was severely deformed under the same conditions.

The lower aluminum-containing alloys displayed excellent wear resistance in metal turning tests. For example, tool inserts of Composition 4 in Table II were used to machine the nickel-base alloy, Ren 41. It was found that the wear rate of the WC-Ni-Al tool was about 40% less than that of the WC-10 volume percent Co tool material used commercially for this purpose.

EXAMPLE 3 WC skeletons having 10 volume percent porosity were prepared by acid-leaching the cobalt from WG-10 volume percent Co composites and subsequently hydrogen-firing, as in Example 1. They were then infiltrated with a series of cobalt-aluminum alloys containing 2.5 to 30% aluminum. About 5% WC was included in the infiltrant when the Al content was less than 15% and 2% carbon was added to all infiltrants. Infiltration was accomplished at various temperatures, depending upon the Al content, as in Example 2. The resulting physical properties are shown in Table III.

The effect of varying aluminum levels on the properties of WC-Co-Al materials is thus seen to be comparable to those obtained with corresponding WC-Ni-Al compositions. Although it would be possible to conventionally sinter a material consisting of WC volume percent Co with up to about 10 weight percent of the matrix being aluminum, great difficulty would be encountered to avoid oxidation of the aluminum present in fine powder form, even when pre-alloyed with cobalt. By way of contrast, there is no aluminum oxide present in the structure of the infiltrated WC-Ni-Al materials.

Even if oxidation of aluminum were no problem, it is not possible to reach full density using liquid-phase sintering when the aluminum content exceeds about 10 weight percent of the matrix, since the solubility of WC in such alloys is very low.

The WC10 volume percent Co-Al alloys were found to possess the same improved high temperature deformation resistance and machining properties that were described in Example 2 for WC-lO volume percent Ni-Al alloys.

EXAMPLE 4 A WC-lO volume percent cobalt composition was acidleached such that the cobalt was removed to a depth of .010 inch below the surface. The porosity network thus generated was then infiltrated with gold by placing the required amount of gold foil over the compact and heating to 1150 C. in a hydrogen atmosphere.

Subsequent corrosion tests in a NaCl acetic acid solution (22 hr. exposure time) showed a weight loss of 2.6 milligrams per square decimeter per day for the gold surface infiltrated sample vs. 258 milligrams per square decimeter per day for the WC-lO volume percent Co control sample, a one hundredfold improvement.

Cutting inserts /2" x /2" x A of initial composition 72% WC, 8% TiC, 11.5% Tac, 8.5% Co were leached for one week in boiling HCl as in Example 1 above, forming carbide skeletons that were 88% dense. These were subsequently infiltrated at 1600 C. with Co-% Al alloy containing additionally 5% WC and 2% carbon. They were then used to machine SAE 1045 teel under conditions which produce significant nose deformation of the original cobalt matrix composition (measured bulge of .005"). Under these conditions the Co-Al matrix material was significantly less deformed (measured bulge of .001").

What is claimed is:

1. A process for producing a cemented carbide composition comprising sintering a pressed mixture of a refractory metal carbide and an iron group binder metal,

removing at least a portion of the binder from the sintered alloy to produce a carbide skeletal structure, and

replacing the removed binder by infiltration into the skeletal structure with a second binder metal different than the Original binder to produce a fully dense cemented carbide composition having a finegrained, homogeneous structure.

2. The process of claim 1 in which substantially all of the binder is removed and replaced by the second binder metal.

3. The process of claim 1 in which the binder is removed by chemically leaching with hot dilute'acid.

4. The process of claim 1 in which the refractory carbide is one or more carbides of tungsten, titanium, or tantalum.

5. The process of claim 4 in which the original binder is less than 20 volume percent of colbalt, nickel, iron or alloys thereof.

6. The process of claim 5 in which the replacement binder is a nickel-base alloy.

7. The process of claim 5 in which the replacement binder is a cobalt-base alloy.

8. The process of claim 5 in which the replacement binder is gold, silver or copper.

9. A process for producing a cemented tungsten carbide composition comprising sintering a pressed mixture containing tungsten carbide and less than 20 volume percent of a cobalt binder, chemically removing the cobalt binder from the sintered composition, and

replacing the removed binder by infiltration into the skeletal structure of a second binder metal other than cobalt to produce a fully dense cemented tungsten carbide composition having a fine-grained, homogeneous structure.

References Cited UNITED STATES PATENTS 2,108,797 2/1938 Comstock -204 2,731,711 1/1956 Lucas 29-1818 3,303,559 2/1967 Holtzclaw 75-208 X OTHER REFERENCES High Temperature Materials-Recent Advances in Infiltrated Titanium Carbide. Lavendel and Goetzel, April 1957, pp. -154.

CARL D. QUARFORTH, Primary Examiner A. J. STEINER, Assistant Examiner U.S. Cl. X.R.