The present invention relates to a cermet material suitable for use as a hardface of a composite roll, a composite roll comprising a metal core and a hardface of a cermet material, and a method of forming a hardface on a metal core.

The present invention relates particularly, although by no means exclusively, to rolls used in the production of structural steel. Such rolls are usually chill cast from steel-based or iron-based materials.

The main requirements of rolls used in the production of structural steel are:

(a) abrasion resistance to minimise wear and localised damage to the roll surface;

(b) capability to wear uniformly so that it is possible to maintain close dimensional tolerances in the rolled product;

(c) resistance to thermal fatigue crack initiation and propagation;

(d) compressive and tensile strength to resist mechanical forces; and

(e) toughness.

The relative importance of these requirements varies with the stages in the rolling mill. For example, resistance to thermal fatigue crack initiation and propagation is a more important requirement in the roughing stand than in the finishing stand. Therefore, a roll which has a combination of properties which is acceptable for use in a finishing stand may not be acceptable for use in a roughing stand.

It has been proposed to use composite rolls comprising a metal core and a hardface of a cermet material in the rolling of structural steel. On a theoretical basis, cermet materials should be well suited to provide the requirements noted above, since by selecting the separate components of a cermet material to comprise a tough matrix and a hard carbide dispersion in the matrix and by adjusting the relative proportions

of the components it should be possible to optimise the properties of the hardface to overcome the problem that some of the requirements noted above, such as abrasion resistance and toughness, are not necessarily compatible. However, on a practical basis, it has hitherto not been possible to produce a cermet material hardface on a metal core to form a roll which adequately complies with the requirements noted above. To a certain extent, this situation has been contributed to by difficulties applying a seemingly suitable cermet material to a metal core to form a hardface which is acceptable in terms of microstructure and bonding to the metal core. The microstructure considerations include the need to have a uniform dispersion of carbide particles and minimal local defects. A further factor which has made it difficult to use cermet materials is that it must be possible to machine the hardface t. form a surface of acceptable quality.

An object of the present invention is to alleviate the disadvantages described in the preceding paragraphs.

According to the present invention there is provided a cermet material suitable for use as a hardface of a composite roll comprising, a dispersion of niobium carbide particles in a cobalt based matrix or a nickel based matrix having a carbon concentration in the matrix of less than 0.3 wt.%.

It is preferred that the matrix carbon concentration is less than 0.25 wt.%.

It is preferred particularly that the matrix carbon concentration is less than 0.22 wt.% for the

cobalt based matrix and less than 0.20 wt.% for the nickel based matrix.

It is preferred that the niobium carbide particles comprise less than 80 vol.% of the cermet material.

It is preferred particularly that the niobium carbide particles comprise less than 50 vol.% of the cermet material.

The term "cobalt based matrix" is herein understood to mean a matrix formed from an alloy which comprises cobalt as the major component. The alloy may include any one or more of the following elements.

(a) Chromium

The alloy may include up to 40 wt.% chromium. Chromium forms a solid solution with cobalt. Chromium carbides form in the presence of carbon.. In addition, chromium oxides form as a surface layer which provides excellent high temperature oxidation and corrosion resistance.

(b) Nickel

The alloy may include up to 25 wt.% nickel. The nickel is added to improve the ductility of the matrix.

(c) Molybdenum

The alloy may include up to 4 wt.% molybdenum. The molybdenum forms carbides and is added to increase the hardness and hot strength of the matrix.

The molybdenum also allows the matrix to age harden.

(d) Tungsten, Niobium, Tantalum

The alloy may include up to 15 wt.% tungsten, up to 6 wt.% niobium and up to 2 wt.% tantalum. These elements have generally the same effect as molybdenum on the properties of the matrix.

(e) Titanium

The alloy may include up to 1 wt.% titanium. The titanium forms an intermetallic compound with nickel which allows the matrix to age harden.

(f) Aluminium

The alloy may include up to 1 wt.% aluminium. The aluminium forms an intermetallic compound with nickel which allows the matrix to age harden.

(g) Silicon

The alloy may include up to 1.5 wt. % silicon. The silicon acts as a deoxidiser.

(h) Manganese

The alloy may include up to 1.5 wt. % manganese. The manganese acts as a deoxidiser and a desulpheriser.

(i) Iron

The alloy may include up to 5 wt.% iron. The iron is used as a cheap alloying element which has no effect on the matrix

when added in amounts less than 5 wt.%.

( ) Vanadium

The alloy may include up to 1 wt.% vanadium. The vanadium forms carbides which increase the hardness of the matrix.

The term "nickel based matrix" is herein understood to mean a matrix formed from an alloy which comprises nickel as the major component. The alloy may include any one or more of the following elements.

(a) Chromium

The alloy may include up to 20 wt.% chromium. Chromium forms a solid solution with nickel at room and at elevated temperatures. The chromium does not form intermetallic compounds and is not subject to phase changes which cause age hardening. Chromium oxides form as a surface layer which provides excellent high temperature oxidation and corrosion resistance.

(b) Iron

The alloy may include up to 10 wt.% iron. In amounts up to 5 wt.% the iron is used as a cheap alloying element with no disadvantageous effects on the matrix. However, higher amounts of iron may reduce corrosion resistance and hot strength of the matrix.

(c) Molybdenum

The alloy may include up to 20 wt.%

molybdenum. The molybdenum improves resistance to acids and increases the hot strength of the matrix.

(d) Cobalt

The alloy may include up to 20 wt.% cobalt. The cobalt forms intermetallic compounds with titanium, molybdenum and aluminium which allows the matrix to age harden.

(e) Tungsten, Niobium, Tantalum

The alloy may include up to 15 wt.% tungsten, up to 6 wt.% niobium and up to 2 wt.% tantalum. These elements form carbides and are added to increase the hardness of the matrix.

(f) Titanium

The alloy may include up to 3 wt.% titanium. The titanium forms intermetallic compounds with tungsten, cobalt and aluminium which allows the matrix to age harden.

(g) Aluminium

The alloy may include up to 6 wt.% aluminium. The aluminium forms intermetallic compounds with cobalt and titanium which allows the matrix to age harde .

(h) Vanadium

The alloy may include up to 1 wt.%

vanadium. The vanadium forms carbides which increase the hardness of the matrix.

(i) Silicon

The alloy may include up to 6 wt.% silicon. The silicon acts as a deoxidiser, a fluxing agent, and improves corrosion resistance of the matrix.

( ) Manganese

The alloy may include up to 3.5 wt.% manganese. The manganese acts as a deoxidiser and desulphuriser and improves hot corrosion resistance of the matrix.

According to the present invention there is also provided a composite roll comprising a steel based or iron based core and a hardfacing of the cermet material described in the preceding paragraphs.

According to the present invention there is also provided a method of forming a composite roll comprising, welding a cermet material comprising particles of niobium carbide and a cobalt or nickel based alloy onto a core of steel based or iron based material to form a hardface comprising a dispersion of niobium carbide particles in a cobalt based matrix or a nickel based matrix having a carbon concentration in the matrix of less than 0.3 wt.%.

It is preferred that the matrix carbon concentration is less than 0.25 wt.%.

It is preferred particularly that the matrix carbon concentration is less than 0.22 wt.% for the

cobalt based matrix and less than 0.20 wt.% for the nickel based matrix.

It is preferred that the niobium carbide particles comprise less than 80 vol.% of the cermet material.

It is preferred particularly that the niobium carbide particles comprise less than 50 vol.% of the cermet material.

It is preferred that the metallic particles are spheroidal and comprise low oxygen and low slag concentrations. It is also preferred that the metallic particles are between 45 and 150 micron in diameter.

It is preferred that the diameters of the niobium carbide particles are between 45 and 90 micron. Alternatively, the diameters of the niobium carbide particles may be within the following ranges:

(a) 5.4 - 22.5 micron;

(b) 22.5 - 45 micron;

(c) 5.6 - 45 micron;

(d) 16 - 63 micron;

(e) 16 - 90 micron; and

(f) 5 - 20 micron.

It is preferred that the method further comprises, welding a buffer material having a matrix carbon concentration of less than 0.3 wt.% onto the core prior to welding the cermet material onto the buffer layer.

One reason for the inclusion of the buffer layer is to avoid an increase of the matrix carbon

concentration of the hardface above 0.3 wt.% due to dilution of the cermet material as a result of melting of the core in the molten pool formed during welding of the cermet material. Another reason for the inclusion of the buffer layer is to avoid an increase in the matrix iron concentration above that present in the cermet material.

It is preferred that the method comprises plasma transferred arc welding of the cermet material and, where applicable, the buffer material onto the core.

It is preferred that the powder and plasma gas comprises between 25 and 100 vol.% argon and up to 75 vol.% helium. In this regard, it has been found that the use of 100 vol.% argon as the powder and plasma gas results in less primary niobium carbide dissolving in the molten pool formed during welding of the cermet material and hence less secondary carbide precipitating. Furthermore, the use of 100 vol.% argon also results in a more uniform carbide distribution. However, it has also been found that the overall cermet material hardness is reduced for a given vol.% of niobium carbide when 100 vol.% argon is used as the powder and plasma gas. It has also been found that the use of helium in the powder and plasma gas has the beneficial effects of reducing nozzle build-up and allowing a thicker weld deposit.

It is preferred that the shielding gas comprises argon 100 vol.%.

It is preferred that the method comprises electromagnetic stirring of the weld pool.

The present invention is based to a large extent

on the realisation that, in order to minimise welding and machinability problems and to provide a hardfacing which has the optimum properties for use in rolling applications, the hardfacing should comprise a cermet material comprising a dispersion of niobium carbide particles in a cobalt or nickel based metallic matrix in which the carbon concentration is less than 0.3 wt.%, preferably less than 0.25 wt.%.

Specifically, it has been found in relation to the significance of matrix carbon concentration that:

(a) cracking and other weld defects increase with matrix carbon concentration of the hardfacing;

(b) a hardfacing microstructure with carbon in solution rather than precipitated as secondary carbides is easier to machine and not significantly lower in abrasion and wear resistance; and

(c) the matrix carbon concentration influences the dispersion of niobium carbide in the matrix such that at low matrix carbon concentrations there is a uniform dispersion of niobium carbide and as the matrix carbon concentration increases there is an increase in the likelihood of segregation of the niobium carbides.

It has also been found that:

(a) the matrix iron concentration has a marked effect on the properties of the hardfacing and it is important to avoid dilution of the cermet material with iron from the core; and

(b) electromagnetic stirring of the weld pool reduces the size of the matrix carbides which should

have a beneficial affect on thermal fatigue crack initiation and propagation and does not adversely affect the dispersion of niobium carbide particles in the matrix.

The present invention is described further by way of example with reference to the accompanying drawings in which:

Fig. 1 is a series of schematic views illustrating the deposition structure of NbC/Nistelle C, NbC/Stellite 21 and NbC/316 stainless steel cermet material hardfacings and buffer layers welded onto steel substrates as part of an experimental procedure to evaluate the present invention;

Fig. 2 is a photomicrograph (magnification 160) of a typical view of a 50 vol.% NbC/Nistelle C hardfacing;

Fig. 3 is a photomicrograph (magnification 160) of a typical view of a 30 vol.% NbC/Nistelle C hardfacing;

Fig. 4 is a photomicrograph (magnification 160) of a typical view of a 20 vol.% NbC/Nistelle C hardfacing;

Fig. 5 is a photomicrograph (magnification 160) of a typical view of a 50 vol.% NbC/Stellite 21 hardfacing;

Fig. 6 is a photomicrograph (magnification 160) of a typical view of a 30 vol.% NbC/Stellite 21 hardfacing;

Fig. 7 is a photomicrograph (magnification 12.8) of a typical view of a 20 vol.% NbC/Stellite 21 hardfacing;

Fig. 8 is a photomicrograph (magnification 160) of a typical view of a 50 vol.% NbC/316 stainless steel hardfacing;

Fig. 9 is a photomicrograph (magnification 12.8) of a typical view of a 30 vol.% NbC/316 stainless steel hardfacing;

Fig. 10 is a photomicrograph (magnification 12.8) of a typical view of a 20 vol.% NbC/316 stainless steel hardfacing; and

Fig. 11 is a plot of Vickers hardness (HV30) versus vol.% NbC for the NbC/Nistelle C, NbC/Stellite 21 and NbC/316 stainless steel hardfacings.

A series of cermet material hardfacings and buffer layers were welded with plasma transferred arc equipment onto 0.2 wt.% steel substrates.

The cermet materials tested comprised:

(a) niobium carbide and Nistelle 21 particles;

(b) niobium carbide and Stellite 21 particles; and

(c) niobium carbide and grade 316 stainless steel particles.

The buffer layer comprised Stellite 21 in the case of the NbC/Stellite 21 and NbC/316 stainless steel

cermet materials and Nistelle C in the case of the NbC/Nistelle C cermet materials.

The chemistries of Stellite 21 and Nistelle C are set out in Table 1 below (amounts in wt.%).

C Co Ni Mn Si Cr M- Fe V B W

Stellite 21 0.2 M

Nistelle C 0.1 - 4.5

316 Stainless Steel 0.02 -

The sketches in Fig. 1 summarise the welding deposition structure for the cermet material hard facing and buffer layers.

The photomicrographs in Figs. 2 to 4 show that the welding of 50 vol.%, 30 vol.% and 20 vol.% NbC/Nistelle C on a Nistelle C buffer layer produced a uniform distribution of primary NbC in a matrix displaying a dendritic dispersion of eutectoid with no weld defects or linear carbides.

The photomicrograph in Fig. 5 shows that the welding of 50 vol.% NbC/Stellite 21 on a Stellite 21 buffer layer produced an uneven distribution of primary NbC particles in a matrix bearing a dendritic dispersion of rather course NbC precipitated from solution and a dendritic dispersion of finer carbides associated with the Stellite 21. There were no weld defects or linear carbides. Nevertheless, the uneven distribution of primary NbC particles is unacceptable from a rolling viewpoint.

The results for the 30 vol.% NbC/Stellite 21 were equally if not more unfavourable. With reference to Fig. 6, the photomicrograph reveals pronounced segregation between primary and larger secondary NbC particles within a matrix bearing fine NbC and Stellite 21 carbides in dendritic arrangement.

The results for the 20 vol.% NbC/Stellite 21 were significantly better than the 50 vol.% and 30 vol.% NbC/Stellite 21 samples. With reference to Fig. 7, the photomicrograph reveals a comparatively even distribution of primary and secondary NbC particles in a matrix bearing fine NbC and Stellite 21 carbides in dendritic arrangement. There were no weld defects or linear carbides.

It has been found that the failure of the 50 vol.% and 30 vol.% NbC/Stellite 21 samples was not due to an inherent problem with the use of a cobalt based matrix but was due to the level of carbon in the matrix being too high. In this regard, it is noted that the concentration of carbon in Stellite 21 is 0.2 wt.% as opposed with only 0.1 wt.% in Nistelle C and that in view of the free carbon in NbC it would be expected that carbon concentration problems would be more evident with Stellite 21 samples than with Nistelle C samples and with increasing concentrations of NbC. On the basis of empirical observations it has been determined that the matrix carbon concentration should not exceed 0.3 wt.% and preferably 0.25 wt.%.

It is noted that the significance of the matrix carbon concentration coupled with the inevitability in practice that some core material will melt into and form part of the hardfacing is an important factor in the need to use a buffer material having a relatively low

carbon concentration. Another contributory factor is that it has been found that an excess of iron in the matrix has an adverse affect and therefore an increase in iron concentration above that present in nickel and cobalt based matrix materials as a result of melting of the core material should be avoided. The adverse affect of matrix iron is reflected in the experimental results for NbC/grade 316 stainless steel.

The photomicrographs in Figs. 8 to 10 show that the welding of 50 vol.%, 30 vol.% and 20 vol.% NbC/316 stainless steel revealed significant segregation of primary and larger secondary NbC particles. The matrix comprises coarse and fine NbC and fine unidentified carbides in a dendritic arrangement. There were no weld defects or linear carbides. Nevertheless, as noted above, the uneven distribution of primary NbC particles is unacceptable from a rolling viewpoint.

With reference to Fig. 11, it is evident that the hardness of the NbC/Nistelle C and NbC/Stellite 21 samples plateaued over the range of 50 vol.% and 30 vol.%. This finding leaves open the possibility that for many applications it would be sufficient to use dispersions of NbC as low as 30 vol.%.

Many modifications may be made to the preferred forms of the invention described above without departing from the spirit and scope of the invention.