Coated Cemented Carbide Product

The outstanding properties of the aluminum oxide-coated product of the invention depend upon careful control of the process parameters. The process involves the use of a gaseous mixture of hydrogen, water, and an aluminum halide such as aluminum trichloride. Carbon monoxide and carbon dioxide may be optionally added. The primary overall deposition reaction is:

The most important ingredients in the gaseous reaction mixture are therefore water vapor and aluminum chloride vapor. However, the aluminum chloride vapor can be formed in several ways during the deposition reaction, as for example by heating solid AlCl.sub.3 powder or by passing chlorine gas over aluminum metal. The water vapor is more conveniently formed by reacting hydrogen with carbon dioxide in the deposition chamber to form carbon monoxide and water vapor by the water gas reactions:

the amount of water vapor formed in this manner depends upon the temperature and the initial concentrations of hydrogen, carbon dioxide, carbon monoxide and water vapor in the input gas stream. In order to form a good quality coating of desirable thickness in the temperature range of 900.degree.-1250.degree.C., the ratio of water to hydrogen gases present, after the water gas reaction, should be between about 0.025 and 2.0.

Hydrogen has been found to be necessary in the vapor deposition process to obtain a dense, adherent coating. Hydrogen appears to insure oxidation of the aluminum at the carbide surface. Oxidation in the reaction zone above the carbide substrate creates a condition known as dusting -- which must be avoided. The absence of hydrogen creates a porous coating which is not fully dense. Thus the three necessary ingredients of the process are aluminum halide vapor, water vapor and hydrogen. In its preferred form, the process includes the use of aluminum chloride vapor, hydrogen and carbon dioxide, the latter reacting with H.sub.2 to form water vapor.

The amount of water vapor present, after the reaction of known input concentrations of H.sub.2 and CO.sub.2, and CO and H.sub.2 O if used, can be calculated using the following equation: ##EQU1## where a = 1 - K:

k = the equilibrium constant for the water gas reaction:

b = (CO).sub.i - (H.sub.2 O).sub.i + K ((H.sub.2).sub.i + (CO.sub.2).sub.i + 2 (H.sub.2 O).sub.i): and

c = K ((H.sub.2).sub.i (CO.sub.2).sub.i + (H.sub.2).sub.i (H.sub.2 O).sub.i + (H.sub.2 O).sub.i (CO.sub.2).sub.i + (H.sub.2 O).sub.i.sup.2).

The parentheses denote the concentration of the gaseous species enclosed within in terms of partial pressure, and the subscripts (f) and (i) denote the final or equilibrium concentrations and the initial or input concentrations, respectively. The amount of H.sub.2 present, and thus the H.sub.2 O/H.sub.2 ratio, can then be determined from the relationship: (H.sub.2).sub.f = (H.sub.2).sub.i + (H.sub.2 O(.sub.f - (H.sub.2 O).sub.f.

A series of coated products wwere prepared accordance with the invention by passing aluminum chloride vapor, hydrogen and carbon dioxide over cemented carbide inserts. The examples were prepared at various input gas compositions and at various final H.sub.2 O/H.sub.2 concentrations. In all cases, deposition was at 1050.degree.C. and a 45-minute deposition cycle was used with 2-3 grams of aluminum chloride and an aluminum chloride generator temperature of about 200.degree.C. The use of more AlCl.sub.3 shifts the desired H.sub.2 O/H.sub.2 ratio to a higher value and vice versa. The coatings were deposited on a cemented carbide substrate having the following composition in % by weight: WC-72, Co-8.5, TiC-8, TaC-11.5. Table I below shows the effect of gas composition on coating thickness. When coating with both higher and lower ratios of H.sub.2 O/H.sub.2 (i.e. outside range of about 0.025 to 2.0), It wasn't possible to get a coating of sufficient thickness, i.e., > 1 micron. Coating quality was good for all examples having more than 1 micron thickness coating. Coating quality was judged to be good if the coating could withstand an adherency test consisting of sliding the coated insert under a diamond brale indentor of the same type used for the Rockwell hardness test using a load of 2 kilograms on the diamond. If the coating resisted spalling or crumbling during this test, it was judged to have good quality. If it did not, it was judged to have poor quality.

The nature of the coating obtained was determined by using X-ray diffraction analyses and optical microscopy. X-ray analyses showed the coating to be alpha Al.sub.2 O.sub.3. At the higher deposition temperatures (greater than 1150.degree.C.), significant amounts of the compound W.sub.3 CO.sub.3 C began to form due to reaction of the substrate carbide with the coating atmosphere. Optical microscopy revealed a gray, translucent coating of Al.sub.2 O.sub.3 that was fully dense and well bonded to the substrate in those examples in which the coating quality was found to be good. A very thin (less than 1 micron) layer of another nonmetallic compound, cobalt aluminate (CoAl.sub.2 O.sub.4), was present between the Al.sub.2 O.sub.3 layer and the cemented carbide substrate. The presence of this thin layer is necessary to achieve the proper bond strength between the coating and the substrate, that is, a bond strength sufficient to pass the adherency test set forth above. In those cases in which no observable intermediate nonmetallic layer was present, the coated inserts did not pass the above described adherency test. For this reason, a cobalt (iron, or nickel) aluminate intermediate layer is believed necessary to a good quality coating.

The preferred temperature range for deposition of the coating is 900.degree.C. to 1250.degree.C. At lower temperatures, the deposition rate becomes very low and the coating is poorly bonded to the substrate. At higher temperatures, excessive reaction occurs between the coating atmosphere and the cemented carbide substrate, weakening the bond between the coating and the substrate and lowering the strength of the overall composite body.

The strength of the Al.sub.2 O.sub.3 coated cemented carbide composite was measured (as were all strength measurements disclosed herein), using a slightly modified standard transverse rupture test (ASTM No. B4066-63T), that included three roll loading and a span-to-thickness ratio of 3.5 to 1. Using a deposition temperature of 1050.degree.C. and a cemented carbide substrate of the nature set forth in the first ten examples in Table I above, the average strength of bars having coating thicknesses of from 5 -7 microns (the preferred thickness for this substrate in terms of wear resistance) was 241,000 psi. This represents only a slight reduction (11 percent) from the 270,000 strength value obtained from the uncoated cemented carbide substrate.

In the following Table II, the metal cutting performance of coated inserts prepared in accordance with this invention is shown and compared with the corresponding performance of uncoated inserts. Examples II through 17 were 1/2 .times. 1/2 .times. 3/16 inches disposable cutting inserts, coated with Al.sub.2 O.sub.3 at 1050.degree.C. by the vapor deposition technique disclosed above for Examples 1 through 10. A range of coating thicknesses of from 1-10 microns was used. These inserts were then used to machine SAE 1045 steel, 190 BHN hardness, at 700-, 1000- and 1500- surface-feet-perminute speeds, 0.010 inches per revolution feed, and 0.100 inch depth of cut. The cutting times to a flank wear of 0.010 inch are shown in Table II, along with the crater wear depth at the 0.010 flank wear time. The transverse rupture strengths are also given. For comparison purposes, the cutting performance and strengths of the uncoated substrate material. Examples 18 and 19, a commercially available solid aluminum oxide base (89% Al.sub.2 O.sub.3. 11% TiO) insert -- Examples 20-22 -- and a Ti C coated cemented carbide insert -- all run under the same conditions -- is also shown in Table II.

It can be seen that the cutting performance of the cemented carbide tool material is very substantially improved by the Al.sub.2 O.sub.3 coating and that this improvement is substantially greater than a TiC coating on the same substrate. It is also evident that the amount of improvement obtained is dependent upon coating thickness up to a value of about 7 microns and that some evidence of performance decline occurs at 10 microns. At the optimum thickness value of 7 micons for this substrate, the performance of the Al.sub.2 O.sub.3 coated tool was equivalent to that of the solid Al.sub.2 O.sub.3 tool at all three speeds tested. The strength of the Al.sub.2 O.sub.3 coated inserts was, however, considerably higher than solid Al.sub.2 O.sub.3 and higher than the strength of the same substrate with a TiC coating.

It should be noted that, because of strength limitations, it has not been feasible to use solid aluminum oxide cutting materials in disposable cutting inserts of the type used in pin-type holders. These inserts have a centrally disposed hole for the reception of a pin which locks the insert in place. The strength of such inserts must be sufficient to resist the locking stresses. The strength of the present coated materials is sufficiently high to enable their use in such inserts. The present invention therefore makes possible the use of an insert, in such applications, having a higher wear resistance than any comparable insert presently available.

The following Table III shows the performance of the coated inserts of the invention in cutting a high temperature nickel-base alloy, specifically Inconel 718 in the solution-aged condition (BHN 390 hardness). The results, Example 25, are compared with the performance of an uncoated cemented carbide of the same composition (Example 26), and in addition with a commercial solid aluminum oxide tool (Example 27). The inserts were of the negative-rake disposable type (indexable and invertible) and were 1/2 .times. 1/2 .times.3/16 inches. The cemented carbide substrate for Examples 25 and 26 was 94% WC and 6% Co. The substrate was coated with Al.sub.2 O.sub.3 at 1050.degree.C. by the vapor deposition process described above in connection with Examples 1 through 10.

The performance of the insert coated with 2.5 microns of Al.sub.2 O.sub.3 was significantly better than that of the uncoated cemented carbide insert of the same substrate composition. From tests with other coating thicknesses, it has been determined that the optimum thickness for this kind of machining (i.e., high temperature alloys) is in the 1-3 micron range. Thicknesses greater than 3 microns in these tests decreased tool life. The superior strength of the Al.sub.2 O.sub.3 coated tools is amply demonstrated by the rapid failure of the solid Al.sub.2 O.sub.3 tool in Example 27, whereas no breakage or chipping was observed in the Al.sub.2 O.sub.3 coated tools, Example 25.

The foregoing is a description of illustrative embodiments of the invention, and it is applicant's intention in the appended claims to cover all forms which fall within the scope of the invention.