The present invention relates to a method of making ce¬ mented carbide articles using binder phase powders with spherical, non-agglomerated particles.

Cemented carbide mainly contains tungsten carbide and cobalt, often along with certain other carbides, e.g. carbides of titanium, tantalum, niobium, chromium etc. It contains at least one hard but brittle (carbide) phase and a relatively less hard but ductile and tough metal, particularly cobalt. This results in materials combining hardness and toughness which have found many applications, for instance in rock drilling and metal cutting tools, in wear parts etc.

Cemented carbide is made by techniques usual in powder metallurgy, that is :

- mixing the constituent powders (carbides, cobalt, and possibly other hard materials) by milling, using mills (rotating ball mills, vibrating mills, attritor mills etc) equipped with non-polluting milling media which themselves are made of cemented carbide. The milling is made in the presence of an organic liquid (for instance ethyl alcohol, acetone, etc) and an organic binder (for instance paraffin, polyethylene glycol, etc) in order to facilitate the subsequent granulation operation.

- granulation of the milled mixture according to known techniques, in particular spray-drying. The suspension containing the powdered materials mixed with the organic liquid and the organic binder is atomised through an ap¬ propriate nozzle in the drying tower where the small drops are instantaneously dried by a stream of hot gas, for instance in a stream of nitrogen. The granules col-

lected at the lower end of the tower have an average di¬ ameter adjustable by the choice of appropriate nozzles, between 100 and 200 μm. Such granules flow easily, in contrast to fine or ultrafine powders. The formation of granules is necessary in particular for the automatic feeding of compacting tools used in the subsequent stage.

- compaction of the granulated powder in a matrix with punches (uniaxial compaction) or in a bag (isostatic compaction) , in order to give the material the shape and dimensions as close as possible (considering the pheno¬ menon of shrinkage) to the dimension wished for the fi¬ nal body. If necessary, the compacted body can be sub- jected to a machining operation before sintering.

- sintering of the compacted bodies at a temperature and during a time sufficient for obtaining dense bodies with a suitable structural homogeneity.

The sintering can equally be carried out at high gas pressure (hot isostatic pressing) , or the sintering can be complemented by a sintering treatment under moderate gas pressure (process generally known as SINTER-HIP) .

The sintered cemented carbides can be characterised in particular by their porosity and their microstructure (observed by optical or electron microscopy) .

The cobalt powders conventionally used in the cemented carbide industry are obtained by calcining cobalt hy¬ droxide or oxalate followed by a reduction of the oxide so obtained by hydrogen; see for instance "Cobalt, its Chemistry, Metallurgy and uses", R.S. Young Ed. Reinhold Publishing Corp. (1960) pages 58-59. These conventional cobalt powders are characterised by a broad particle

size distribution, with strongly aggregated particles, in the form of agglomerates with a sponge-like aspect, which are difficult to mill since there are strong binding forces between the elementary particles in these aggregates.

In EP-B-0113281, the making of metallic powders with a process for reducing oxides, hydroxides or metal salts with the aid of polyols, is described. Particularly when starting with cobalt hydroxide, it is possible to obtain powders of metallic cobalt with essentially spherical and non-agglomerated particles. Further studies have shown in particular that it is possible to obtain non- agglomerated metallic powders having controlled average diameters of the particles, for instance by varying the concentration of the starting hydroxide or metal salt, in relation to the polyol(s). It is in this way that, in the case of cobalt, it is possible to obtain particles with an average diameter of, for instance 1, 2 or 3 μm, by using the ratios cobalt hydroxide/polyol of 0.033, 0.1 or 0.340 g cobalt/c ^ polyol, respectively. Simi¬ larly, it is possible to obtain particles with adjust¬ able average dimensions, smaller than 1 μ by seeding the reaction mixture with the aid of very fine metallic particles (for instance palladium) either by adding a metal salt or hydroxide reacting more quickly than the cobalt salt or hydroxide with the polyol. This is parti¬ cularly the case with silver salts, in particular silver nitrate, which are quickly reduced to metallic silver in the form of very fine particles of which the number is roughly proportional to the quantity of silver introduc¬ ed into the reaction chamber. The silver or palladium particles so formed serve as seed for the growth of co¬ balt particles which are subsequently formed by reduc- tion of the cobalt hydroxide or salt by the polyol. The higher the number of seed particles, the smaller the di-

mensions of the final cobalt particles. For instance, when using a molar ratio silver/cobalt in the range of 10- -10 ^~ 2, _{one can} obtain cobalt particles having average dimensions that vary from 0.1 to 0.3 μ and the range can be extended by varying this ratio between 10 ^" ^ and 10 ^~ 1 for all the appropriate metals. These various methods for controlling the the size of the metallic particles are particularly known and described by M. FIG ARZ et al. M.R.S.Int'l Mtg on Adv. Mats. Vol 3, Ma- terials Research Society, pp. 125-140 (1989), F.FIEVET et al. Solid State Ionics 32/33, 198-205 (1989) and F. FIEVET et al. MRS Bulletin, December 1989, pp.29-34.

It has now been discovered that the cobalt powders ha- ving the properties of those obtained by the reduction of cobalt hydroxide or a cobalt salt with the aid of po¬ lyol, according to the EP-B-0 113 281 and the references just mentioned, that is powders of individual, essenti¬ ally spherical non-agglomerated particles, can be used as binder phase powder in the manufacture of cemented carbide and that this preparation gives several advanta¬ ges which are exposed below.

It has been particularly discovered that when using such non-agglomerated cobalt powders, it is possible to ob¬ tain in a reproducible way, cemented carbide exhibiting interesting characteristics, in particular, a reduced porosity. It is also possible to decrease the milling time for starting mixtures (carbide and binder) without impairing the quality of the final cemented carbide. Acceptable results can be obtained even after a simple blending operation. Alternatively, the degree of milling may be further reduced and the cemented carbide subjected to a hot isostatic pressing process, either incorporated into the sintering process or as a separate operation, giving an increase in the grain size of the

hard phase and correspondingly an increase in resistance to thermal cracking.

In addition, it has been discovered that, due to the use of such cobalt powders, it is possible to carry out the sintering at temperatures below those which are gene¬ rally used. This decrease of sintering temperature is interesting not only from an energy point of view, but also because it permits the possibility of adding to the powder mixture other hard or superhard materials (in the form of powders) which cannot normally be used at the temperature required for conventional sintering. Among these other superhard materials, one can note particu¬ larly diamond, of which it is known that it starts transforming into graphite in air at a temperature around 800 °C and cubic boron nitride. Alternatively, the sintering temperature may be lowered even further and the cemented carbide subjected to a hot isostatic pressing process, either incorporated into the sintering process or as a separate operation, giving an increased hardness level and a more uniform grain size and binder phase distribution leading to an increase in mechanical strength.

Generally, the cobalt particles used as binder phase ac¬ cording to the invention have dimensions that can vary from 0.1 to 20 μm, in particular from 0.1 to 10 μm. Par¬ ticles having dimensions from 0.1 to 5 μm are particu¬ larly used. Especially interesting results have been ob- tained with submicron particles (that is with a size •less than 1 μm) .

The present invention has thus as object the use - as binder phase, in the preparation of cemented carbide by milling, then sintering a mixture of powders with at le¬ ast one hard material based on tungsten carbide and a

binder phase - of at least one powder of cobalt, that is homogeneous as regards the size of the particles, and particularly one powder having an average particle size x (in for instance the range 0.1 to 20 μm, in particular between 0.1 and 10 μ ) , of which at least 80% of the particles have sizes in the interval x ±0.2x, provided the interval of variation (that is 0.4x) is not smaller than 0.1 μ . The cobalt powder used in accordance with the invention consists of individual, essentially sphe- rical and nonagglomerated particles.

Such powders can be especially obtained with the polyol reduction which is recalled below. It is preferable to start with cobalt hydroxide or cobalt acetate.

The cobalt powders obtained by the reduction of cobalt hydroxide with the aid of polyol generally contain a small proportion of carbon (most often less than 1.5% by weight) and oxygen (most often less than 2.5% by weight) . These powders can be directly used in the manu¬ facture of cemented carbides.

Generally, according to the invention the cobalt powder used as binder in the preparation of cemented carbide will exclusively be a powder such as defined above. But it is possible to use such powders in combination with a second cobalt powder exhibiting other characteristics, provided the proportion of the first powder is suffici¬ ent for giving the advantages indicated in the prepara- tion of cemented carbide, for instance a decrease of the sintering temperature. Generally the first powder repre¬ sents at least 10%, and preferably at least 50% of the total weight of the cobalt used as binder phase.

In addition it is possible to use as binder phase a mix¬ ture of two or more powders as defined above, these two

powders having different average particle dimensions.

It has also been found that the use of cobalt according to the invention is very suitable to adjust the binder content of an already dried cemented carbide mixture. Such an adjustment is not possible with a conventional binder phase powder since the resulting mixture lacks necessary flowability. Adding polyol cobalt does not adversely affect flowability and can even improve it. Thus, a unique 'mother-mix' may be used for producing a variety of cemented carbide grades having different binder phase contents. After the addition of the polyol- cobalt, preferably with a particle size of <3 μ , up to the desired content the mixture is ho ogeneized before pressing and sintering.

The starting powder mixture contains cobalt in suffici¬ ent proportions for the final cemented carbide to con¬ tain 0.1 to 40% by weight of cobalt, and preferably 3 to 25 %. It is particularly advantageous in grades with very low contents of cobalt (typically <0.5%) often re¬ ferred to as binderless grades.

Sintered cemented carbide bodies based on WC, particu- larly with a grain size <1.5 μ , manufactured according to the method of the invention has a porosity better than A02 and BOO, less than 0.5, preferably less than 0.2, binder phase lakes per cm^ with a maximum dimension of >25 μm and less than five carbide grains per cm^ with a grain size of more than 5 times the average grain size of the matrix.

In the manufacture of cemented carbides where the sin¬ tered grain size of the hard phases is fine, i.e. 1 μm or less, it is commonplace to substitute a small amount of other refractory metal carbides for tungsten carbide.

The carbides commonly used are those of titanium, tanta¬ lum, niobium, vanadium, chromium and hafnium. The effect of these substitutions is to control grain growth of the hard phase during sintering. A side effect is that they inhibit melt formation during sintering with the result that often higher sintering temperatures are needed than would be the case without the substitution to ensure freedom from microporosity and a uniform binder phase (cobalt-rich phase) distribution. The result is to partly negate the advantage of the substitution, leading to a degree of grain growth, recrystallization, of the C-phase which results in a nonuniform grain size dis¬ tribution, a less than optimum hardness level and a re¬ duction in mechanical strength. Using the cobalt-powder according to the invention the above-mentioned grain growth inhibitors may be excluded. This applies in par¬ ticular to high pressure anvils for diamond production in which the cobalt-content of the cemented carbide is 5-7 weight-% and WC grains size <1.5 μm. Another example is tools such as drills, microdrills and routers for ma¬ chining of printed circuit boards and similar composite materials. Such tools have a cobalt-content of 3-20 weight-%, preferably 4-12 weight-% and a WC grain size of <1 μ , preferably <0.7 μm.

For certain applications where a degree of thermal shock is experienced, for example hot rolling of steel bar, some mining and highway engineering applications and ma¬ chining of stainless steel, it is desired that the hard phases should be of relatively coarse grain size, typi¬ cally greater than 4 μ preferably greater than 6 μm and the cobalt content <10, preferably <8 weight-%. A ce¬ mented carbide powder to produce such a sintered hard phase grain size must of necessity be relatively lightly milled in order to control the degree of comminution. The result is that the degree of intimate mixing is re-

duced, and, owing to the coarse particle size, the area available for reaction during sintering to produce a melt is relatively small. Consequently, such cemented carbide powders prove to be difficult to sinter and re- quire high temperatures to approach a fully dense condi¬ tion. Using the non-agglomerated, spherical cobalt pow¬ der dense bodies can be obtained at a lower sintering temperature.

In EP-A-0182759 it has been shown that an increased strength was obtained in sintered bodies of cemented carbide being used in tools for rock drilling. The but¬ tons according to this patent has a core consisting of a fine-grained eta-phase M5C (e.g. C03W3C) and/or M12C (e.g. CogWgC) , embedded in normal alpha (WC) and beta

(Co binder phase) structure at the same time as the sin¬ tered body has a surrounding surface zone which consists of alpha and beta-phase in two areas whereas the outer shell is cobalt depleted and the inner part has a high content of binder phase. Surprisingly it has now been found that cemented carbide bodies manufactured in such a way as described above give a more optimized toughness behaviour when cobalt according to the invention is used in the production of the buttons. The effect is most ut- ilized for cemented carbide with a cobalt content of more than 10% and less than 25% by weight and preferably 13% to 20% by weight of cobalt. The mean grain size of the hard constituents is larger than 1,5 μ . The same appearance has also been obtained for cemented carbide bodies with mean grain size of alpha-phase (WC) of less than 1,2 μm and a binder content of equal or less than 6% by weight of cobalt. When cobalt according to the in¬ vention is used in the sintering/heat treatment proce¬ dure the sintering temperature can be reduced which re- suits in a lower carbon content in the binder-phase and a low porosity level. The benefit of this sintering/heat

treatment gives a product with a high carbon activity and a fine grain size eta-phase which results in a ce¬ mented carbide body with a more pronounced difference in cobalt content in the surface zone between the outer co- bait depleted shell and the inner part rich with cobalt. The cemented carbide produced with the cobalt according to the invention has a cobalt content with greater dif¬ ference and reduced width of the shells in the surface zone which leads to higher compressive stresses in the surface zone and has also positive effects on strength and toughness.

The invention has been described above with reference to the manufacture of conventional cemented carbide i.e. based upon WC and with a binder phase of cobalt. It is evident that the invention also can be applied to the manufacture of articles of other composite materials with hard constituents (borides, carbides, nitrides, carbo-nitrides etc) and a binder phase, based on cobalt, nickel and/or iron, such as titanium based carbonitride alloys usually named cermets. Said alloys are manufactu¬ red by milling powders of carbides, nitrides and/or car- bonitrides of mainly Ti but also of other metals from groups Via, Va and Via of the periodical table of the elements (V, Zr, Nb, Mo, Ta, W etc) together with pow¬ ders of nickel and cobalt. The mixture is then dried, pressed and sintered as described above for conventional cemented carbide.

Example 1

A suspension of cobalt hydroxide was put in a mixture of monoethylenglycol and diethylenglycol, while agitating. The suspension, containing about 200 g of cobalt hydrox- ide per liter, was progressively heated to a temperature of at least 200 °C, while strongly agitating. A solution

of silver nitrate was then added in the monoethyleng- lycol, so that between 0.07 and 0.3 g silver per liter was introduced. The mixture was kept at the same tempe¬ rature during 2 hours, and was then left to cool to room temperature.

In this way a cobalt powder (reference PI) was obtained with the following properties:

- SEM diameter of the particles: 0.4 μ - C : 1.36% by weight

- 0 : 2.23% by weight

The SEM diameter is the average diameter of the par¬ ticles measured in the scanning electron microscope.

In addition the following raw materials were used:

Tungsten carbide :

- Origin : Eurotungstene Poudres (France) - Total carbon : 6.15% by weight

- Free carbon : 0.05% by weight

- Average diameter (Fisher) : 0.9 μm

Tantalum carbide : - Origin : H.C.STARC

- Total carbon : 6.81% by weight

- Free carbon : 0.10% by weight

- Niobium : 9.09% by weight

Cobalt (reference F) obtained by reduction of the oxide with hydrogen according to the conventional process:

- Origin : Eurotungstene Poudres

- Diameter according to Fisher : 1.30 μ

- C : 0.012% by weight

With the aid of these materials the following mixtures

were prepared:

- Cobalt: 3% or 6.5% by weight

- Tantalum carbide: 0.5% by weight

- Tungsten carbide balance

The powder mixture (500 g) was obtained by milling in a mill of the type "Attritor" with a capacity of 9 liter, containing 3.5 kg of milling media (balls of cemented carbide with a diameter of 3 mm) turning at 250 turns/minutes, in the presence of 200 ml of ethyl alco¬ hol (or acetone) and with the addition of polyethylene glycol (2 g per 100 g of mixture) . The powder was milled during 7 or 14 hours and thereafter granulated using a sieve with 120 μ mesh size. The compaction was carried out under uniaxial compaction from two directions, with matrix and punches of cemented carbide under a pressure of 125 MPa. Sintering was performed at 1375, 1410 and 1450 °C respectively. After sintering microsections were prepared and the porosity and recrystallisation were de- termined.

The porosity was determined according to the standard ISO 4505 and is expressed with the aid of a scale of in¬ creasing porosity from A00 to A08.

The recrystallisation of tungsten carbide (or general grain growth) was determined by microscopic examination and visual comparison with an internal standard scale (analogous to that of the ISO scale for the porosity) since no standard exists up to this day. The results are expressed with a scale going from Rl (quasi-absence of recrystallisation) to R5 (very strong recrystallisa¬ tion) .

a) -Cobalt : 6.5% by weight -Milling : 14 hours -Sintering : 1450 °C

b) -Cobalt : 6.5% by weight -Milling : 7 hours -Sintering : 1450 °C

c) -Cobalt : 3% by weight -Milling : 14 hours -Sintering : 1375, 1410, or 1450 °C -Results before HIP :

Results :

Sintering temperature, °C 1375 Type of cobalt PI F Porosity A02 A08

Cobalt lakes* s N * The average number of cobalt lakes was determined by counting (in an optical microscope) the lakes on ten op¬ tical fields at a magnification of 1500 times and taking the average. s : a few N : Numerous

d) -Cobalt ; 3% by weight -Milling : 14 hours -Results after HIP

The HIP treatment consists in putting the samples sin¬ tered during the previous experiment in a HIP furnace at 1350 °C, during 2 hours, under 100 MPa (atmosphere : ar¬ gon) .

Result :

Sintering temperature, °C 1375 1410 1450 Type of cobalt Pi F PI F PI F

Porosity A01 A01 A01 A01 A01 A01

Cobalt lakes s N 0 N 0 N s : a few

N : numerous 0 : no

These tests show clearly, that all other factors equal, the use of cobalt according to the present invention shows itself as beneficial in comparison with a conventional cobalt since it entails a decrease of porosity and of the number of cobalt lakes.

Example 2

Two laboratory scale batches of cemented carbide powder were made using the same batch of tungsten carbide, this batch having an average particle size of about 1 μ as measured by the Fisher sub-sieve sizer method. In grade A 6% by weight of conventional hydrogen-reduced cobalt powder was added and in grade B 6% by weight of ultra- fine spherical cobalt powder was added. The same small addition of chromium carbide powder was added to each grade. A fairly intense degree of milling was given to each grade by milling 1 kg of powder with 15 kg of mil-

ling bodies in a liquid for 13.5 hours in a rotary mill. Compacts were made from the dried cemented carbide pow¬ ders and sintered, in close proximity with each other, under vacuum at a range of temperatures. Following sin- tering, microsections were prepared and the porosity le¬ vels were assessed by comparison with standard mi¬ crographs according to method ISO 4505. The binder phase distribution was assessed by an arbitrary method. The specimens were first etched for 4 minutes at room tempe- rature in Murakami's reagent and examined under an opti¬ cal microscope at a magnification of 1500X. The average number of "cobalt lakes" present in a field of view was assessed by counting the number observed in 10 fields and dividing the total count by 10. Cobalt lakes are re- gions of binder phase, typically from 2-10 μ in diame¬ ter, which occur when the sintering temperature was in¬ adequate. The results obtained were as follows:

irosity Co lakes per field

0

0

4.9

0 >200 5.6

From the above results it can be seen that the use of ultrafine spherical cobalt powder in grade B had a marked effect on the level of microporosity and binder phase distribution, especially apparent at the lowest sintering temperature employed. As well as permitting a lower sintering temperature to be employed, the use of ultra-fine spherical cobalt powder confers an improved degree of tolerance to temperature variations within the sintering furnace.

Example 3

Two laboratory scale batches of cemented carbide powder were made using the same batch of tungsten carbide. This batch having a particle size of about 40 μ according to the Fisher subsieve sizer method. The true particle size was however approximately 15 μm, the higher Fisher value being due to agglomeration. In grade C 6% by weight of conventional cobalt powder was added and in grade D 6% of ultra-fine spherical cobalt powder was added. No other carbides were added. A 1 kg charge of cemented carbide powder was milled with 5 kg of milling bodies and a liquid for 13.5 hours in a rotary mill. Compacts were made from the dried cemented carbide powders and sintered, in close proximity to each other, under vacuum at a range of temperatures. Following sintering micro¬ sections were prepared and the porosity levels assessed according to the method detailed in ISO 4505. The re- suits obtained were as follows:

Grade

C D C D C D C D

The above results illustrate that using ultra-fine sphe¬ rical cobalt powder a marked reduction in porosity le- vels was achieved. Thus, lower sintering temperatures may be employed and again an improved degree of tole-

ranee to temperature variation within a furnace change is conferred.

Example 4

Anvils for the 60 mm diamond production system have been tested according to the performance represented as life length in a diamond production. The anvils were manu¬ factured in three different grades of cemented carbide and marked with random numbers prior to the testing. The performance test was applied in a diamond production plant during "normal" working conditions whereas the re¬ sults were reported with life lengths in comparison to presently used anvils. All anvils have a core consisting of a small amount (2%) of eta-phase in the structure.

The anvils of grade A were manufactured according to the conventional production route of cemented carbide and were used as a reference in the test. The anvils were produced as described in example 1 with 6% by weight of conventional hydrogen-reduced cobalt and a small addi¬ tion of chromium carbide. The sintering temperature was 1450 °C and the cemented carbide had a microporosity of A02. The microstructure did not show any cobalt lakes.

The anvils of grade B had a similar composition as des¬ cribed for anvils of grade A without the chromium car¬ bide content. The anvils were subjected to a hot isosta¬ tic pressing process at 4 MPa and 1410 °C instead of standard sintering. No microporosity was obtained in the microstructure and 5,2 cobalt lakes per field were pre¬ sented from microscopic examination of the cemented car¬ bide. The microstructure was even and no influence of discontinuous or local grain growth could be seen.

The anvils of grade C had a composition according to the

invention as described in Example 2 without the chromium carbide content. The anvils were subjected to a hot iso¬ static pressing procedure with the same conditions as for the anvils of grade B. The microstructure examina- tion of the cemented carbide did not show any microporo¬ sity (A00) or cobalt lakes. The structure was even without any influence of discontinuous grain growth.

The α-phase (WC) in the microstructure of the three gra- des of anvils had a mean grain size of about 1,2 μm.

The performance results were reported in actual number of pressings per anvil and scaled in a performance ran¬ king. Each cemented carbide grade was represented by six anvils.

Performance/Rank

D E F A C F C

C D D C B

£ C

Grade C: 1 702 (still in use)B

2 1399 A

3 608 B 4 592 C

5 820 B

6 _2J2£ A Average 837 B

The results of grade A were uneven and the anvils with the low numbers of pressings had cracks in the top of the anvils. Grade B had a better performance but got the same ranking level as grade A. Three anvils had small cracks in the top surface. Grade C had the best perfor- mance ranking in the test and the best pressing beha¬ viour of all anvils. Obviously the anvils according to the invention had the most optimized hardness and tough¬ ness behaviour due to a well dispersed cemented carbide matrix and a narrow grain size distribution of α-phase.

Example 5

A coarse grained tungsten carbide with a grain size of 18 μm was in the as supplied state used to produce test batches of very coarse cemented carbide for concrete and asphalt cutting tools. Cemented carbide with low cobalt content and very coarse grain size is needed to achieve optimum combination of toughness to wear resistance pro¬ perties together with maximum thermal fatigue crack re- sistance.

The same procedure as in Example 3 was used except for that the milling time was reduced to 9,5 hours.

Grade X was produced with 6% of conventional cobalt and grade Y with 0.3 μm of ultrafine spherical cobalt pow-

der. Sintering was performed at 1520 °C in vacuum. Grade X showed a porosity level of A06, BO6 plus 8 pores of 25 μ , and had to be HIP'ed. Grade Y was fully dense with maximum porosity of A02, due to the effective and uni¬ form reduction of the WC-grains together with excellent mixing of the spherical cobalt with the tungsten carbide grains.

The metallographical analysis showed as follows

Grade X Grade Y

Grain size mean value 7 + 4 μm 7 + 1,5 μm maximum size 18 μm 10 μ minimum size 1.8 μm 5 μm structure: uneven with 10-15 even cobalt lakes of 10-20 μ hardness, (HV3) : 1215 1205

Road planing tips were made from the two test batches and were compared with a conventional grade, Z with 8 w/o Co, 5 μm WC grains and a hardness of 1200 HV3. Point attack tools from the three grades were made, and they were geometrically identical with the carbide tips (09 mm, length 18 mm with a conical top) brazed at the same time.

The test was made in hard concrete with an Arrow CP 2000 road planing machine.

Drum diameter: 1 , drum width; 2.2 m

Toolpick speed: 2.0 m/s, cutting depth: 25 mm 180 tools, 60 per grade, were evenly distributed throughout the drum.

Test result (mean value of 50 tools)

Grade Wear, Fractured Rank mm height carbide reduction (no of pieces)

X 5,3 8 2

Y 4,8 1 1

Z 8,1 7 3

Example 6

Buttons for roller bits with diameter 12 mm having a multiphase structure were produced from a small produc¬ tion batch. The average particle size of the WC was 3.5 μm and the nominal cobalt content was 13.5% by weight. The added cobalt was ultra-fine spherical cobalt powder with a Fisher grain size of 0.3 μm. Compacts of the ce¬ mented carbide powder were sintered at 1340°C. Corre¬ sponding buttons were produced with the same production process parameters except of the sintering temperature which was 1380°C. These buttons originating from a ce¬ mented carbide powder blending with conventional cobalt powder with a Fisher grain size of 1.4 μ . All buttons were thermally treated in a carburizing atmosphere for 2 hours. In the following examination of the microstruc¬ ture of buttons from the two batches it could be seen a multiphase structure with a core that contained eta- phase surrounded by a surface zone of cemented carbide free of eta-phase having a low content of cobalt at the surface and a higher content of cobalt next to the eta- phase "core" .

Microprobe studies of the microsections gave the follo¬ wing results:

Grade A (with ultra-fine cobalt) :

Etaphase core (05.0 mm) : mean grain size of etaphase: 4,1 μm mean cobalt content: 11.5 weight-%

Cobalt "rich" zone (width 1,5 mm) mean cobalt content: 14.2 weight-%

Cobalt "depleted" zone (width 2.0 mm) mean cobalt content: 10.0 weight-%

Grade B (according to prior art with conventional co¬ balt) Eta phase core (07.0 mm) : mean grain size of eta-phase: 4,8 μ mean cobalt content: 11.5 weight-% Cobalt "rich" zone (width 1 mm) mean cobalt content: 15.3 weight-% Cobalt "depleted" zone (width 1,5 mm) mean cobalt content: 8.7 weight-%

No porosity could be seen in the surface zone. It is ob¬ vious that buttons prepared according to the invention gave a more distinct multi-phase structure with a higher cobalt-gradient in the surface zone.

Example 7

Wear and toughness tests were performed with roller bits in an open-cut copper mine. The roller bits were of type 9 7/8" CS consisting of three roller cones with spheri¬ cal buttons. The diameter of the buttons was 12 mm. For one roller bit the buttons according to the invention were placed in all positions of the buttons in row 1. Three types of roller bits were used in the test.

Bit A Buttons according to Example 6 were placed as

above and in the excepted positions comparative buttons with the same composition according to prior art .

Bit B Comparative buttons of Example 6 according to prior art in all positions.

Bit C Standard cemented carbide with the same composi¬ tion as in Example 6 but being free of eta-phase and without the multi-phase structure.

Drill rig: 1 pee. BE 45R Feed: 0-60000 lbs Rpm: 60-85 Hole depth: 18-20 m

Type of rock: Biotite gneiss, mica schist.

Resμl ?;

Grade drilled meters index drilling index depth (m/h)

A 1900 160 18 140

B 1650 140 16 120 prior art

C 1170 100 13 100 prior art

The grade according to the invention has obtained longer life length as well as greater drilling rate.

The wear of the buttons was measured at 800 drilled me¬ ters .

Results

Grade A: Row 1: Buttons according to the invention

Average wear 3.0 mm

Row 2 : Average wear 2.8 mm Row 3 : Average wear 2.5 mm

The wear profile gave a self-sharpening effect due to a wear looking like "egg shells". This effect was most marked at row 1. One button was missing in row 1.

Grade B: Row 1: Average wear 3.2 mm

Row 2: Average wear 2.8 mm Row 3: Average wear 2.4 mm

The wear of the buttons was of "egg shells"-type. From row 1 three buttons from one roller cone and two respec¬ tively one from the other two were missing. Two buttons were missing in row 2.

Grade C: Row 1: Average wear 3.6 mm Row 2: Average wear 3.0 mm

Row 3: Average wear 2.6 mm From row 1 five buttons from one roller cone and four respectively one from the other two were missing. The penetration rate was slow at 800 drilled meters.

This test gave surprisingly good results for the roller bit attached with buttons made according to the inven¬ tion. The penetration of the roller bit was also very good.

Example 8

From a 91.5 : 8.5 WC(2 μm)/Co(1.2 μm) powder mixture, granules (hereafter referred to as basic granules) were prepared according to the conventional technique. Then a sufficient amount of cobalt (polyol-type 1 μm) was added to the granules until the respective proportions of WC/Co reached 88:12. After mixing for 30 minutes in a Turbula-type mixer, the resulting mixture ( 'modified granules' ) was tested for flowability according to ISO 4490 with the following results:

Time/100 g, s Basic granules 53

Modified granules 46

After compaction and sintering a cemented carbide was prepared with the basic granules and the modified granu¬ les. The Vickers hardness was determined with the follo¬ wing result:

HV50 Basic granules 1455

Modified granules 1300

As expected the hardness of the cemented carbide with the modified granules was lower than that of the basic cemented carbide in view of the higher cobalt content. The structure of the carbide obtained with the modified granules was satisfactory.