Method of preparing a powder, powder and process for additive manufacturing

Abstract

A method of preparing a metal matrix composite powder for additive manufacturing, by mixing a metal powder with a ceramic powder, binding the mixed powders together to form an electrode; and using the electrode in an electrode induction gas atomisation (EIGA) process to prepare the metal matrix composite powder. The ceramic can be TiB2, SiC, AlN or a MAX-phase ceramic, the latter having the formula Mn+1AXn where M is a metal, A is an A-group element and X is C, N and / or B, e.g. Ti3SiC2, Al4SiC4 or Al8SiC7. The metal powder can be steel (e.g. stainless steel), titanium, aluminium, cobalt, magnesium, copper, refractory metals, PGMs vanadium and alloys thereof (e.g. 6061 aluminium). Binding to form the electrode can be by cold isostatic pressing, sintering and / or the application of pressure. The powder can be processed by laser powder bed fusion AM or direct energy deposition printing.

Description

The present invention relates to a powder for additive manufacturing and in particular to a metal matrix powder for additive manufacturing and a method of preparing such a metal matrix composite powder for additive manufacturing. The invention further relates to a method of manufacturing a metal matrix composite component and an AM component. Metal matrix composites (MMCs) which contain a combination of a metal matrix in which ceramic particles are embedded are attractive for manufacturing components as they can provide a unique balance of properties. For example, in aluminium MMC’s a high level of specific strength is achievable through increased stiffening by the ceramic particles with little density change of the metal. In traditional methods of MMC production such as casting or forming, it is hard to get an even distribution of ceramic in the metal matrix leading to inhomogeneity in physical properties. Also the amount of post machining required to get to the final product form has significant expense. This is because most traditional MMCs contain very hard phases such as SiC, TiBs or other boride, carbide, nitride and oxide phases, so that the MMCs are difficult to machine and require specialized and expensive tooling. The present invention is directed to enabling MMC articles to be manufactured through additive manufacturing techniques. Although manufacturing the article by additive manufacturing methods reduces the amount of post machining necessary, the need for post machining is not necessarily eliminated. It is therefore desirable to develop MMCs that are machinable and which cause reduced tool wear. It is desirable to provide a method of manufacturing MMC components with homogeneous distribution of ceramic and / or in which the amount of post-machining required to get to the final product is low and / or in which post machining is minimised. The present invention provides a method of manufacturing a metal matrix composite component, the method comprising: mixing a metal powder with a ceramic powder; 1 binding the mixed metal powder and ceramic powder together to form an electrode; and using the electrode in an electrode induction gas atomisation process to prepare a metal matrix composite powder, and manufacturing a metal matrix composite component using the metal matrix composite powder in a metal additive manufacturing process. This process has the advantage of producing a powder which has high circularity and including ceramic powder within the metal matrix of the powder particles meaning powder handling during additive manufacturing is improved and the ceramic particles are uniformly distributed in the microstructure of the component. In an embodiment the metal additive manufacturing process is a powder-bed metal additive manufacturing process, preferably a laser powder-bed fusion additive manufacturing process, or wherein the metal additive manufacturing process is direct energy deposition printing. The processes work well with the powder produced by the method. In an embodiment the ceramic powder is TiB2 or SiC or AIN ceramic powder. These powders are suitable for strengthening MMCs. In an embodiment the ceramic powder is a MAX phase ceramic powder. Such powder is suitable for strengthening MMCs and also has the advantage of resulting in a component which is easily machinable. The present invention provides a method of preparing a metal matrix composite powder for additive manufacturing, the method comprising: mixing a metal powder with a MAX phase ceramic powder; binding the mixed metal powder and MAX phase ceramic powder together to form an electrode; and using the electrode in an electrode induction gas atomisation process to prepare a metal matrix composite powder. The powder resulting from such a method has good circulating meaning it is suitable for additive manufacturing and the MAX phase ceramic particles mean the resulting product of AM has good mechanical properties whilst also being machinable relatively easily. In an embodiment the MAX phase of the MAX phase ceramic has a layered hexagonal structure, preferably a layered hexagonal crystallographic structure with weak interlayer bonding and / or wherein the layered hexagonal crystallographic structure is a space group Pbs / mmc. Such structures machine particularly well. In an embodiment the MAX phase of the MAX phase ceramic has a general formula Mn+iAXn, where M is one or more metals, A is one or more A-group element and X is C and / or N and / or B. Preferably M is one or more of: Ti, Sc, V, Cr, Zr, Nb, Hf, Mn, Mo and / or A is one or more of: Cd, Si, Sn, In, Ga, Tl, Al, Ga, Sn, Pb, Zn, Au. Such MAX phases are particularly suited to the task. In an embodiment the MAX phase of the MAX phase ceramic is TisSiC?. In another embodiment the MAX phase of the MAX phase ceramic is A14SiC4 and AhSiC?. In an embodiment the binding step comprises cold isostatic pressing. This ensures that the electrode has enough strength for the EIGA process. In an embodiment the binding step comprises sintering and / or the application of pressure. This ensures that the electrode has enough strength for the EIGA process. In an embodiment the method further comprises sieving the metal matrix composite powder to achieve a predetermined particle size distribution. This ensures uniformity of the component made in the AM process and easier additive manufacturability. In an embodiment the volume of ceramic powder is at least 5% of the total volume of the sum of metal powder and ceramic powder, preferably wherein the volume of ceramic powder is at least 10% of the total volume of the sum of metal powder and ceramic powder, optionally wherein the volume of ceramic powder is 50% or less of the total volume of the sum of metal powder and ceramic powder. Such large quantities of ceramic in MMCs are difficult to manufacture in any other way. In an embodiment the metal powder is a powder of one or more of: nickel and nickel alloys, steels including stainless steels, titanium and titanium alloys, aluminium and aluminium alloys, cobalt and cobalt alloys, magnesium and magnesium alloys, copper and copper alloys, refractory metals and refractory metal alloys, platinum group metals and their alloys, vanadium metals and their alloys. These metals and alloys benefit in particular to having ceramic particles distributed in them to form a MMC. In an embodiment the ceramic has a melting point in Kelvin at least 20% higher that the melting point in Kelvin of the metal of the metal powder. This ensures wider process parameters during the EIGA process as melting the metal can be accomplishes without (undesirably) melting the ceramic. In an embosdiment the metal matrix composite powder has the tenth percentile diameter (dio) between 10 pm and 25 pm and the ninetieth percentile diameter (d9o) between 40 pm and 90 pm. This particle sie is well suited to AM. In an embodiment the metal powder has a dio of less than 15 pm and a dgo of 80 pm or less. This distribution has been found to work well in the method. In an embodiment the ceramic powder has a dio of 5 pm or less and a doo of 10 pm or less. This distribution is well suited for a ceramic phase in a MMC. In an embodiment the metal matrix composite powder has a circularity of 0.7 or more. This level of circularity has been found to be sufficient to allow the powder to be efficiently handled in AM. In an embodiment the mixing includes ball milling, preferably low energy ball milling. Such milling is sufficient to achieve the desired properties of the powder prior to electrode formation and at low energy cost. The present invention provides a metal matrix composite powder comprising: metal matrix composite particles, wherein each metal matrix composite particle comprises a metal matrix in which a plurality of MAX phase ceramic particles are embedded. Such a powder can be used in an AM process to form an article which retains the benefits of a traditional MMC whilst being more easily machinable. In an embodiment the metal matrix composite powder has a circularity of 0.7 or more, preferably of 0.8 or greater. Such a powder is optimised for easy handling in an AM machine due to improved flowability. In an embodiment a volume fraction of ceramic in is in the range of from 5 to 50%. Such a powder can be used to make a product by AM in which the volume fraction of ceramic is higher than is possible with other techniques. In an embodiment metal matrix composite powder has the tenth percentile diameter (dio) between 10 pm and 25 pm and the ninetieth percentile diameter (dw) between 40 pm and 90 pm. Such a powder is suitable for LPBF AM processes. In an embodiment the ceramic particles have a dio of 5 pm or less and a dgo of 10 pm or less. Such a size distribution will substantially remain after AM and makes a suitable reinforcement in a MMC product. In an embodiment the metal consists of 0-10% in sum of silicon, copper, iron, manganese, zinc, titanium, magnesium, lithium, zirconium, beryllium, scandium, sodium, cerium, yttrium, lithium, calcium, silver and the balance being aluminium and inevitable impurities. This is a suitable level of these alloying elements for the metal matrix of a MMC. In an embodiment the MAX phase of the MAX phase ceramic has a layered hexagonal structure. Such structures machine particularly well. In an embodiment the MAX phase of the MAX phase ceramic has a general formula Mn+1 AXn, where M is one or more metals, A is one or more A-group element and X is C and / or N and / or B. In an embodiment M is one or more of: Ti, Sc, V, Cr, Zr, Nb, Hf, Mn, Mo and / or A is one or more of: Cd, Si, Sn, In, Ga, Tl, Al, Ga, Sn, Pb, Zn, Au. Such MAX phases are suitable for use in a MMC component and result in an easily machined component. In an embodiment the MAX phase of the MAX phase ceramic is TisSiC?. In an embodiment the MAX phase of the MAX phase ceramic is A14SiC4 and AhSiC?. The present invention provides an AM component consisting of secondary phase particles of a ceramic powder distributed in an alloy matrix. This is the result of the inventive method. In an embodiment the ceramic particles of the AM component exhibit limited clustering. This is advantageous in terms of achieving homogenous properties. In an embodiment the grain structure of the metal matrix of the AM component is isotropic and monomodal or consists of finer grains near the edges of melt pool boundaries optionally with larger, optionally elongated, grains oriented normally to the melt pool boundaries. This microstructure results from AM manufacturing and achieves good physical properties. The present invention results in a powder which flows easily and results in very good homogeneity of ceramic distribution. Furthermore, due to using a MAX phase as the strengthening component of the MMC, machinability of the eventual MMC article is greatly improved. The near net shape nature of the additive manufacturing process means there is little or no machining required which results in a significant cost reduction. In an embodiment, the present invention relates to a method of preparing an electrode for use in an electrode induction gas atomisation process, the method comprising: mixing a metal powder of with a ceramic powder; and binding the metal and ceramic powders together to form an electrode. The invention will now be described, by way of non-limiting example only, with reference to the following drawings, in which: Figure 1 is a schematic of blended powder. Figure 2 illustrates the difficulty of additive manufacturing with blended metal powder and ceramic powder. Figure 3 is a schematic showing the difference between ball milling and powder blending. Figure 4 are SEM images of AI2O3 and aluminium ball milled powders. Figure 5 shows lack of consolidation in an electrode pressed Figure 6 shows surface defects (right) and cracking (left) in a pressed electrode Figure 7 is a schematic and a photograph of the EIGA process. Figure 8 shows images of electrodes with what is considered a good melt tip (left) and a bad melt tip (right). Figure 9 shows images of incomplete electrode tip detachment during EIGA (left) and fully detached tips (right), as a result of incorrect atomisation parameters Figure 10 are SEM images of powder of Example 1 Figure 11 shows the LPBF processing map of Example I. Figure 12 shows the LPBF processing map of Example 2. Figure 13 is an optical micrograph of the microstructure of Example 1. Figure 14 is a comparison in machinability between Examples 1 and 2. Figure 15 is a comparison in strength values between Examples and benchmark alloys. Figure 16 is a comparison in elongation between Examples and benchmark alloys. Figure 17 is a comparison in stiffness between Examples and benchmark alloys. Figure 18 is a comparison in density between Examples and benchmark alloys. Figure 19 are EBSD images of the microstructure of Example 1. The traditional methods of making a powder for production of MMC materials are the following processes: • Powder blending: mixing a ceramic and metallic powder where the ceramic powder is in contact with the metallic powders but not inside the metallic powder • Ball milling: Mixing metallic constituents and ceramic constituents in a vessel with ball milling media to form a powder which contains ceramic phase within the metal particles • In situ ceramic formation: selecting an alloy composition which precipitates a ceramic phase during solidification Powder blending is the least expensive method to produce a MMC powder. However, it has a number of drawbacks when applied in additive manufacturing processes. First, the ceramic particles are required to be substantially smaller than the metal particles, approximately one order of magnitude smaller. The fine ceramic powder significantly decreases flowability the powder. As fraction of ceramic increases the flowability decreases further increasing the difficulty to additively manufacture components in an MMC with a high volume faction of ceramic particles. Second, the mixed blended powder is agitated throughout use in additive manufacturing. As there is no strong bonding 7 between mixtures as there is a size difference between the ceramic and metal particles it is possible that the agitation of the powder may lead to separation over time changing the actual ceramic content in the blend relative to the target content. Figure 1 shows a blended powder used in a powder bed fusion additive manufacturing method. In such a method components are built up by spreading a layer of powder out and fusing particles of powder together using a laser in areas where the component will end up, before providing a further layer of powder over the powder bed as well as the fused part of the formed component. Areas where this layer should be present are then fused with the laser before continuing with the process to build up the next layer. Figure 2 shows a caking effect which results when blended powders of separate metal particles and ceramic particles are used in such a process. Not only is the powder difficult to handle in terms of flowability over the powder bed (due to the difference in sizes between the metal particles and the ceramic particles and due to the usual non-spherical shape of the ceramic particles), but also separation between the metal particles and the ceramic particles (seen as waves in figure 2) appear. The uneven ceramic distribution in the powder layer is then replicated in the component and the caking resulting from poor flowability hampers efficient manufacturing. Ball milling is a process that includes plastic deformation, welding and fracturing mechanisms. In this technique metal and ceramic powder particles are prepared by ball collisions. From that very fine particle sizes and uniform distribution of the ceramic particles can be obtained (see figure 3). However it has a number of drawbacks when the powder is directly applied in additive manufacturing processes: first, it is difficult to control the resulting powder morphology. The method deforms metal ceramic composite powder, therefore after the process the powder shape is irregular and far from spherical. This irregular powder shape inhibits powder flow significantly. See figure 4 of aluminium / AI2O3 composites formed in this way (from Han, Setchi, and Evans. "Synthesis and characterisation of advanced ball-milled A1-A12O3 nanocomposites for selective laser melting." Powder Technology 297 (2016): 183-192.). Second, the high amount of energy and time required for the ball milling process makes it very expensive. Also the lack of control over resulting powder size makes it difficult to achieve a high yield for powders within a targeted size distribution. The description above refers to high energy ball milling (HEBM). In-situ formation: in some instances an alloy composition can be defined such that the chemistry is completely liquid during melting and during a solidification process the phase transformations in the alloy result in a metallic material with a low ceramic fraction. This can be used to form low volume fraction MMC alloys. However this process is limited to low ceramic volume factions and only to certain metal systems. The present invention has been made to address many of the limitations of the above described traditional methods of making a powder. In particular, a metal matrix composite powder for use in additive manufacturing has been developed along with a method of preparing such a powder. The inventors have conceived of the idea of a metal matrix composite powder in which metal matrix composite particles are formed each of which comprise a metal matrix in which a plurality of ceramic particles are embedded and in which the particles are substantially spherical in morphology. Such a powder addresses deficiencies in flowability (due to the spherical shape) and homogeneity because the particles are formed to have a metal matrix and a plurality of ceramic particles embedded therein. Even uneven powder coverage across the layer of powder on the powder bed during additive manufacturing does not result in inhomogeneous distribution of ceramic particles contrary to the powder blending described above and shown in figures 1 and 2. This is because any separation due to size differences between particles does not lead to segregation of metal and ceramic because the metal and ceramic are present in each particle. Finally, the ceramic used in the MMC is a MAX phase ceramic which, as described below, results in an MMC with high strength properties, yet good machinability. The advantages can be summarized as follows: • Spherical powders are formed so that issues of powder flow are optimized for additive manufacturing processes • The ceramic is completed embedded in the metal particle so that it cannot separate, meaning the faction of ceramic in the powder is controlled through powder processing by additive manufacturing • A high faction of ceramic particle can be included because the process starts from a blended ceramic / powder mixture • The product resulting from additively manufacturing from the powder is more machinable than MMCs using non-MAX phase ceramics The inventors have been able to produce such powder for the first time using electrode induction gas atomization of an electrode which is prepared by mixing a metal powder with a MAX phase ceramic powder, binding the mixed metal and ceramic powders together to form an electrode and using that electrode in an electrode induction gas atomization process to prepare the metal matrix composite powder. Electrode induction gas atomization (EIGA) automatically results in nearly spherical particles being produced. The nature of the electrode naturally results in powders which are comprised of a metal matrix in which MAX phase ceramic particles are embedded. This invention describes a novel gas atomization process to manufacture (optionally high volume fraction) MMC alloy powder, a novel powder for additive manufacture where each powder particle has a reinforcement phase embedded within the metal, unlike blended powder where the metal and ceramic are two distinctly different powders, and additively manufactured materials and component produced by the novel powder. Figure 5 shows scanning electron microscope images of such a powder. Embedding the ceramic within the metal, or for those ceramic particles on the surface of a particle, fixing them to the particle, results in reduced spatter during AM processes where melting occurs. Spatter is where turbulence in the melting pool ejects molten or partly-melted material which lands on the powder bed. These ejecta are generally large and irregularly shaped, which can create defects in the component. Embedding the ceramic within the metal also results in powder with good flow properties which is key to certain AM techniques, especially laser powder bed fusion. Detailed description of embodiments of the process The metallic powder and the ceramic particle are mixed in the desired ratio. The size distribution of the ceramic powder is selected for the desired particle size distribution in the resulting composite. In an embodiment the metal powder has a dio of less than 15 pm (say between 4 and 12 pm , preferably between 5 and 10 pm) and a dgo of 80 pm or less (say between 40 and 80 pm , preferably between 45 and 70 pm) and in an embodiment the ceramic powder has a dio of 5 pm or less (say between 0.1 and 4 pm , preferably between 0.3 and 1.5 pm) and a dgo of 10 pm or less (say between 3 and 10 pm , preferably between 4 and 8 pm). The mixed powders are then bound together to form an electrode. The binding process can include cold isostatic pressing the mixture into the form of an electrode for electrode induction gas atomization (EIGA) process. The electrode is usually cylindrical. The schematic of the EIGA process is given in Figure 6. The binding process can optionally include electrode sintering: Application of temperature and pressure to the cold pressed electrode to increase mechanical strength and density of the electrode prior to EIGA process. The MMC powder is then formed by melting and gas atomization of the electrode using an EIGA atomization process. During this process parameters are varied such that the metal in the electrode is melted but the ceramic remains in substantially the form it was in the original ceramic powder. In one embodiment sieving of atomized particles is carried out to screen powder to an intended size distribution, for example, 15-100 pm or 15-60 pm mean equivalent spherical diameter for laser powder bed fusion additive manufacturing. For example a suitable powder has the following size distribution. Fart / cte size <10 pm <15 pm >53 pm > S3 pm Max 3 5 f Method; by fasertWacito? by sieving ASW 8224 For direct energy deposition (DED) printing, a mean equivalent spherical diameter of 50-100 pm is suitable. This process is particularly suited to producing powders for forming AM compounds with a high (greater than 5% or even 10%) volume fraction of ceramic. Such a high volume fraction of ceramic cannot be achieved by in-situ ceramic formation. Increasing the volume fraction of ceramic too far can result in poor mechanical properties of the finished product (particularly ductility), so in one embodiment the volume fraction of ceramic is limited to 50% or less or even 40% or less. Possible materials The process is suitable for use with a wide variety of metals which can be used in additive manufacturing. A non-exhaustive list includes nickel and nickel alloys, steels including stainless steels, titanium and titanium alloys, aluminium and aluminium alloys, cobalt and cobalt alloys, magnesium and magnesium alloys, copper and copper alloys, refractory metals and refractory metal alloys, platinum group metals and their alloys, vanadium metals and their alloys. In this context alloy means a mixture of the main metal with up to 70 wt% of alloying elements (though for non-superalloys and refractory alloys a maximum of 50 wt% of alloying elements may be more appropriate). EIGA is a flexible process by which one can produce powders of all the above alloy systems suitable for powder metallurgy processes. EIGA is particularly well-suited to alloys that cannot be produced via crucible gas atomization due to the limitations of the refractory ceramics used in crucibles (e.g. refractory alloys) or the difficulty achieving a homogeneous and low-viscosity melt (e.g. metal matrix composites as described in this invention). A particularly promising metal for additive manufacturing is aluminium and aluminium alloys. In particular, an aluminium alloy consisting of 0-10% in sum of silicon, copper, iron, manganese, zinc, titanium, magnesium, lithium, zirconium, beryllium, scandium, sodium, cerium, yttrium, lithium, calcium, silver and the balance being aluminium and inevitable impurities. Ceramics which have higher melting point than the metal with which they are combined are suitable as the ceramic. The melting point of the ceramic (both measured in Kelvin) is preferably at least 20% higher than that of the metal matrix. It is also preferable that the reactivity of the ceramic with the metal is low. Low reactivity means the ceramic particles will largely retain their original phase composition and fraction after EIGA and the additive manufacturing process - a small degree of reactivity however is acceptable. For example, up to 20 volume percent of the original ceramic phase may react with the metal matrix. The degree of reaction can be measured by scanning electron microscopy (SEM) techniques and image processing. One example of this is to take a suitably flat and polished sample of the additively manufactured material described by this invention and acquiring EDX and EBSD scans of at least 100 microns by 100 microns. One can then use standard SEM software for automatic phase analysis that makes use of composition data 12 from EDX and crystallographic data from EBSD and matches it to an internal database to identify the phases in the scan. Note that the ceramic phase may partly dissolve in the metal matrix during atomisation and additive manufacturing but reprecipitate as fine particles of the same phase during solidification and heat treatment. This is not considered as reactivity. A particularly promising ceramic powder is titanium diboride (TiB2) which is particularly suitable as a ceramic in a metal matrix composite. In particular a combination of titanium diboride and aluminium or aluminium alloy forms a suitable MMC powder. This is in fact the powder illustrated in figure 5. Another particularly promising ceramic powder is AIN which is particularly suitable as a ceramic in a metal matrix composite. In particular a combination of AIN and aluminium or aluminium alloy forms a suitable MMC powder. Another particularly promising ceramic powder is SiC which is particularly suitable as a ceramic in a metal matrix composite. In particular a combination of SiC and aluminium or aluminium alloy forms a suitable MMC powder. TiB2, TiC, AIN and SiC have the disadvantage that they are very hard and so result in low machinability of the produced article. This is acceptable in the case where the product can be additively manufactured to the exact final dimensions or very close thereto. However where this is not possible there are advantages in using a ceramic which results in better machinability of the 3D printed MMC product. Particularly promising ceramic powders are MAX phases. MAX phases are characterized by their high ratio of specific stiffness to hardness and a layered hexagonal structure (space group Pba / mmc) with weak bonds between the layers, which can easily delaminate. MAX phases typically consist of at least 3 elements, 2 of which are transition metals or metalloids, while the third is typically either C, N or B (or a mixture thereof). A general formula is often Mn+iAXn, where M is one or more metals, particularly an early transition metal from group 3 to group 7 of the periodic table and A is one or more A-group (mostly IIIA and IVA or groups 13 and 14) element and X is C and / or N and / or B. Known examples of M are Ti, Sc, V, Cr, Zr, Nb, Hf, Mn, Mo but also Al. Known examples of A are Cd, Si, Sn, In, Ga, Tl, Al, Ga, Sn, Pb, Zn, Au, S, P. In an embodiment n=l-7. Note that these rules cover most, though not all MAX phases - again, their key characteristic is the layered hexagonal crystallographic structure (typically space group P63 / mmc) with weak interlayer bonding. The low hardness of MAX phases and their ability to delaminate makes them much easier to machine than ceramics typically used in MMCs (TiB2, SiC, TiC). MAX phases are preferred for making machinable MMCs. A particularly promising ceramic powder is MAX phase Ti3SiC2 which is particularly suitable as a ceramic in a metal matrix composite. In particular a combination of Ti3SiC2 and aluminium or aluminium alloy or a steel forms a suitable MMC powder. Another particularly promising ceramic powder are Al4SiC4 and AhSiC? which are particularly suitable as ceramics in a metal matrix composite. In particular, a combination of A14SiC4 and AhSiC? and steel or aluminium or aluminium alloy forms a suitable MMC powder. Other examples of MAX phases are included in the following table: A list of example MAX phases 211 Ti2CdC, Sc2InC, Sc2SnC,Ti2AlC, Ti2GaC, Ti2InC, Ti2TlC, V2A1C, V2GaC, Cr2GaC, Ti2AlN, Ti2GaN, Ti2InN, V2GaN, Cr2GaN, Ti2GeC, Ti2SnC, Ti2PbC, V2GeC, Cr2AlC, Cr2GeC, V2PC, V2AsC, Ti2SC, Zr2InC, Zr2TlC, Nb2A1C, Nb2GaC, Nb2InC, Mo2GaC, Zr2InN, Zr2TlN, Zr2SnC, Zr2PbC, Nb2SnC, Nb2PC, Nb2AsC, Zr2SC, Nb2SC, Hf2InC, Hf2TlC, Ta2AlC, Ta2GaC, Hf2SnC, Hf2PbC, Hf2SnN, Hf2SC, Zr2AlC, Ti2ZnC, Ti2ZnN, V2ZnC, Nb2CuC, Mn2GaC, Mo2AuC, Ti2AuN 312 Ti3AlC2, Ti3GaC2, Ti3InC2, V3A1C2, ThSiCz, Ti3GeC2, Ti3SnC2, Ta3AlC2, Ti3ZnC2, Zr3AlC2 413 Ti4AlN3, V4A1C3, Ti4GaC3, Ti4SiC3, Ti4GeC3, Nb4AlC3, Ta4AlC3, (Mo,V)4A1C3 514 Mo5VA1C4 Electrode formation and atomisation A metal powder and a ceramic powder are first mixed. Generally, the ceramic powder will have a mean particle diameter of 0.05 - 500 pm which is the desired mean particle diameter (equivalent spherical diameter) in the finished product. The size distribution is usually measured by different techniques such as laser diffraction analysis (according to ASTM B822), dynamic light scattering (according to ASTM B822), and electron microscopy. In one embodiment the ceramic powder has a unimodal size distribution with a dio diameter of 0.3-2 pm and a d9o diameter of 3.0-9.0 pm measured by ASTM C1070. The mean particle diameter of the metal powder is not so critical. This is because the metal melts and is reshaped into a different particle size distribution during the electrode induction gas atomization process. However, if the metal powder is too large, this can result in non-uniform mixing of the metal powder and ceramic powder resulting in the reduced consolidation of the electrode. Moreover this can result in the inhomogeneity of the final atomised powder. The metal powder will have a dio diameter of 10-20 pm and a dgo diameter of 50-70 pm measured by ASTM B822. Metal powder that contains more fines than specified can result in the agglomeration of the ceramic particles during ball milling and cold isostatic pressing. It has been found that using non-spherical metal and ceramic particles is advantageous as it results in better binding of the metal and ceramic powders during cold isostatic pressing. In an embodiment, the circularity of the metal powder is less than 0.8 and preferably less than 0.6. For the ceramic particles, the circularity is less than 0.5 and preferably less than 0.3. Circularity is a measure of how closely a particle resembles a circle and is defined below. The metal powder and ceramic particles may be blended through any suitable or desirable method, preferably drum blending. The blended material is then processed using a low energy ball mill in a dry state at variable speeds and duration (e.g., 10 - 50 rpm for 25 minutes). Optionally high energy ball milling may be used. It has been found that using low energy ball milling during blending of the metal powder with the ceramic powder is advantageous. This is because the purpose of this step is primarily the homogeneous blending of the metal powder and the ceramic particles and not powder attrition. Advantages of using low energy ball milling (LEBM) compared to high energy ball milling (HEBM) in the blending of the metal powder and ceramic particles 15 include but are not limited to lower temperature rise during the milling process, controlled particle size reduction of the final milled powder, a lower degree of agglomeration, and a lower degree of powder morphology transformation. For one embodiment of this invention with a matrix of A16061 (a precipitation hardened aluminium alloy with 0.80-1.20 wt% Mg, 0.40-0.80 wt% Si, 0.15-0.40 wt% Cu and 0.04-0.35 wt% Cr with optional additions of Fe up to 0.70 wt%, Zn up to 0.25 wt%, Ti up to 0.15 wt% and Mn up to 0.015 wt% with other elements present in an amount of 0.05 wt% or less each with a total sum of other elements of 0.15 wt% and incidental impurities) and volume fraction of 15% TnSiCT ceramic, the step of ball milling comprises: placing the pre-blending powder in the drum ball mill, milling at a speed of for example 50 rpm for at least 25 minutes using a ball to material ratio of 10:1. Following the blending of the metal powder and the ceramic powder, the powder mix is bound together to form an electrode. The binding may comprise cold isostatic pressing (CIP) and optionally may include sintering. For CIP, wet or dry pressing can be used in obtaining the green body. The pressure applied is defined based on the matrix material and could vary between 50 - 600 MPa and preferably 200 - 450 MPa for a minimum duration of 10 seconds and preferably more than 2 minutes. This will result in a consolidated rod suitable for the sintering step. If inappropriate parameters are selected, this can lead to poor consolidation of the electrode (as seen in Figure 5) resulting in surface defects and cracking (as seen in Figure 6) The sintering condition depends on the matrix materials; the sintering temperature is defined by the liquidus and solidus of the matrix and for A16061 could vary between 300 -650 °C and preferably 500 - 620 °C. As a general rule, the sintering temperature should lie in the mushy zone (the coexistence of at least one liquid and at least one solid phase) of the matrix alloy. This will result in optimum electrode density. A lower temperature, at least 0.7Tm where Tm is the liquidus of the alloy in Kelvin, works too but may require a longer time to achieve good consolidation. The duration of the process should preferably be more than 10 minutes and preferably more than 1 hour. The parameters and in particular the current, voltage, feed rate, and axisymmetric rotation speed used in the electrode induction gas atomization process (see Figure 7) are adjusted dependent upon the unique combination of matrix metal and ceramic reinforcement the 16 electrode is made of. These parameters will determine the formation of a melt film and subsequent stability of the molten metal stream or droplets that fall from the electrode tip (as seen in Figure 8) into the inert gas nozzle to form spherical-shaped fine powder. The current and voltage values are determined using the following rules where the electrode temperature can be increased by increasing current or voltage. If the electrode is made too hot, this can result in the vaporisation of the metal and / or the generation of high thermal gradient from the melt tip into the body of the electrode resulting in melt tip discontinuation and eventual tip detachment (as seen in Figure 9). However if the electrode is not hot enough, the flow of metal stream is poor caused by a partially formed melt tip ( as seen in Figure 8). Poor flow may result in a preferential melting and atomisation of the matrix alloy and the agglomeration of ceramic reinforcement in the tip. In these conditions, obtained atomised powder particles have reduced ceramic content while ceramic atomises as agglomerates of much larger size. These agglomerates are preferably sieved out for the powder to conform to the AM process requirements. As a result an MMC powder with reduced ceramic fraction is produced. This is not desirable. Although the EIGA process parameters are dependent on the setup (shape of induction coils, gas nozzle, and other key parts which tend to vary between machines) general conditions are necessary for obtaining powders suitable for AM. These include atomising the electrode bars with diameters between 30 to 100 mm using an inert gas, preferably argon, for the melting and atomisation processes. In this invention, we found that an axisymmetric electrode rotation speed of between 1 to 10 RPM with a feeding velocity of between 10 mm / min to 60 mm / min is preferred to achieve a more uniform electrode melting. A supply pressure of the atomisation gas is preferably between 18-25 bar and a frequency of the resonant circuit powering the induction coils is between 105 and 125 kHz depending on the atomisation setup. As mentioned earlier, control of the continuous metal melt and the PSD of the atomised powder is specific to the EIGA setup used and processing parameters need to be optimised accordingly. The particles atomized in the EIGA process can be sorted (e.g. sieved) to select a suitable (predetermined range) size for laser powder bed fusion additive manufacturing. A suitable size distribution is such that its dio is larger than 20 pm and dgo is smaller than 85 pm as measured by ASTM B822. In an embodiment, the metal matrix composite powder consists of 10 vol% or less of particles with a size of <15 pm according to ASTM B822 and 10 vol% or less with a size of >100 pm according to ASTM B214. Furthermore, when plotting the size distribution of the powder based on the two methods or combinations thereof, the distribution is ideally unimodal (has a single maximum). In an embodiment, the metal matrix composite powder consists of 5 wt% or less of particles with a size of <10 pm, 10 wt% or less with a size of <15 pm, both according to ASTM B822 and 20 wt% or less with a size of >53 pm and 5 wt% or less with a size of >63 pm both according to ASTM B214. Furthermore, when plotting the size distribution of the powder based on the two methods or combinations thereof, the distribution is ideally unimodal (has a single maximum). [One suitable size distribution is given in the below table: Partite size <10 pm <15 pm >53 pm > S3 pm yax 3 U® 5 K® f w® Method: by / asertfOwtto? ASIM BS22 by sfewng ASW 8214 In Example 1 of this invention, the initial mix of raw powders consists of 15 vol% TiB2 ceramic phase containing only inevitable impurities with dio of 0.5 pm to 1.2 pm and dgo of 4 pm to 7 pm and the remainder of aluminium alloy 6061 according to the specification, with dio of >7 pm and dgo of <60 pm. The resultant powder after the milling, CIP and EIGA steps retains the initial distribution of TiB2 embedded in 6061 particles of high circularity with dio of 12 pm to 20 pm and dgo of 40 pm to 100 pm. Figure 10 shows the microstructure of the powder of Example 1. A desirable, spherical powder morphology with embedded, evenly dispersed TiB2 particles is clearly visible. The target circularity is above 0.8 (measured as described below). In Example 2, the initial mix of raw powders consists of 10 vol% Ti3SiC2 MAX phase ceramic powder containing only inevitable impurities with dio of 10 pm and d9o of 60 pm and the remainder of aluminium alloy 6061 according to the specification, with dio of >7 pm and d9o of <60 pm. The resultant powder after the milling results powder particles of Ti3SiC2 embedded in 6061 alloy matrix. The powder particles have a dio of 12 pm to 20 18 pm and dw of 40 pm to 100 pm. The size distribution of embedded TisSiC? particles is somewhat broader than initially and skewed towards smaller size due to its attrition during the milling process. The MMC powder is preferably suitable for additive manufacturing (AM), with a narrow, microscopic size distribution, lack of impurities, good circularity and flowability. The preferred powder morphology and size distribution depends on the exact AM method used. The preferred AM method in this invention is laser powder bed fusion (LPBF), for which the preferred size distribution is discussed above. In an embodiment, when analysing the cumulative distribution function of particle diameters, the tenth percentile diameter (dio) is preferably between 10 pm and 35 pm and the ninetieth percentile diameter (dgo) is preferably between 50 pm and 100 pm. The diameter is measured using laser diffraction either from the volume equivalent of a sphere or the area equivalent of a circle. The average aspect ratio of the powder, defined in the usual way as the ratio of the longest Feret diameter of a particle to the shortest one, is preferably below 2. The average circularity of the powder is preferably between 0.7 or greater, more preferably of 0.8 or more and 1.0 and less and is defined by the formula 4ttA / P2 where A is the projected area of the powder particle and P is the overall perimeter of the projection, according to ISO9276-6 . Aspect ratio and circularity are measured on images taken using optical microscopy or scanning electron microscopy on a sample of powder deposited on a flat surface. The sample should be thin enough to form a single layer of particles in which particles have limited overlap and minimal vertical stacking. The preferred flow properties of the powder as measured by Powder flow results from Carney and Hall flow methods should be less than 11 s / 50 g and 35 s / 50 g (the time it takes for 50 g of powder to pass through the funnel with a circular orifice diameter D = 0.508 cm and 0.254 cm respectively), respectively. Additive manufacturing process The mixed powder can then be manufactured into an item using an additive manufacturing process. This involves depositing and melting a layer of powder, allowing it to solidify, depositing and melting another layer of powder, allowing it to solidify and repeating the process, adjusting the cross-section and thickness of the layer as well as the laser scan parameters as necessary, to produce the target item. Alternatively, the additive 19 manufacturing process can involve depositing multiple layers of powder mixed with a binding agent. The layers can again differ in cross-sections, thicknesses and the parameters used for deposition. This route needs to end with a sintering step that volatilises the binder phase and allows the MMC powder particles to sinter. The additive manufacturing process is preferably a mature process such as directed energy deposition (DED) or laser powder bed fusion (LPBF). Unexpectedly it has been found that the addition of ceramic particles, which remain solid in the melt of the matrix alloy, increase the viscosity of the melt pool, which results in a stable, viscous, less turbulent melt pool, which leads to a wider processing window, as well as reduced defect density and as-built roughness. The stable melt pool also contributes to improved feature fidelity compared to using matrix alloy powder alone. This increased feature fidelity is essential in printing complex, intricate geometries. Ceramic particles with good wettability also provide desirable properties by acting as heterogeneous nucleation sites in the melt, resulting in finer grains. Fine grain size is beneficial because it increases ductility and strength and reduces the propensity to hot cracking. The preferred method of AM in this invention is laser powder bed fusion (LPBF). The preferred LPBF process parameters depend on the particular combination of matrix material and ceramic reinforcement. In an embodiment, the process preferably should result in a printed article with an optical density >98%. Optical density is measured by first acquiring an optical or scanning electron micrograph of a polished part of the cross-section of the article, then using image analysis techniques to quantify the area associated to porosity and finally dividing it by the total measured are to derive optical density. For Example 1 of this invention, the preferred LPBF process parameters are: • A volumetric energy density between 40 J mm’3 and 120 J mm’3 to achieve optical density of >99.7% perpendicular and parallel to build direction. The laser speed and power can vary between 800 mm s’1 and 1450 mm s’1, as well as between 200 W and 325 W, respectively. Preferably laser speeds and powers of 800 mm s’1 to 1000 mm s’1 and 200 W to 250 W is ideal. • A hatch spacing between 0.07 mm and 0.11 mm. Preferably a hatch spacing of 0.075 mm to 0.085 mm is ideal. • A layer thickness between 20 pm and 90 pm. Preferably a layer thickness of 20 pm to 40 pm is ideal. • A laser focus diameter between 0.04 mm and 0.12 mm. Preferably a laser focus diameter of 0.05 mm to 0.07 mm is ideal. Figure 11 shows the optical density of items produced using the process in Example 1. It shows that a broad range of 3d energy density values (also known as volumetric energy density or VED) can result in very low porosity in the microstructure (Porosity = 100% -optical density [%]) indicating a broad processing window. VED is a metric commonly used to compare different processing parameters. The formula for calculating VED (square brackets indicate the unit used in calculations) is: VED [J mm’3] Laser power [W] Powder layer thickness [mm] x Hatch spacing [mm] x Laser velocity [mm s-1] Figure 13 is an optical micrograph of the microstructure of Example 1, showing a favourable homogeneous dispersion of TiB? ceramic particles. Figure 15 shows a comparison in machinability between Example 1 and 2. Samples of both materials were turned on a lathe with a surface speed of approximately 60 m s'1 using an uncoated cemented tungsten carbide tool. Example 1 wears down the tool rapidly, resulting in the rapid rise in workpiece roughness, and indicating the need for harder cutting tools, such as those with diamond-like carbon coatings or tools made of boron nitride or diamond. For Example 2 of this invention, the preferred LPBF process parameters are: • A volumetric energy density between 40 J mm’3 and 120 J mm’3 to achieve optical density of >99.7% perpendicular and parallel to build direction. The laser speed and power can vary between 800 mm s’1 and 1450 mm s’1, as well as between 200 W and 325 W, respectively. Preferably laser speeds and powers of 1150 mm s’1 to 1350 mm s’1 and 200 W to 250 W is ideal. • A hatch spacing between 0.07 mm and 0.11 mm. Preferably a hatch spacing of 0.085 mm to 0.105 mm is ideal. • A layer thickness between 20 pm and 90 pm. Preferably a layer thickness of 20 pm to 40 pm is ideal. • A laser focus diameter between 0.04 mm and 0.12 mm. Preferably a laser focus diameter of 0.05 mm to 0.07 mm is ideal. Examples 1 and 2 have a microstructure consisting of secondary phase particles of a ceramic powder dispersed in an alloy matrix. In can be seen that the ceramic powder is evenly distributed in the alloy matrix exhibiting limited clustering. The ceramic particles have a dio of 5 pm or less and a d9o of 10 pm or less and comprise 5 vol% to 50 vol% of the microstructure of the component. Preferably, examples of the invention exhibit limited ceramic clustering and uniform volume / area fraction throughout the component (as is the case for Example 1). Preferably, the grain size of the alloy matrix has a narrow unimodal distribution, preferably with a dio of 2 pm or less and a d9o of 7 pm or less (as is the case for Example 1). EBSD maps in Figure 19 show this microstructure - the larger grains are uniformly dispersed TiB2 while the finer grains are the aluminium alloy matrix. It is apparent that the latter are considerably finer than TiB2. The grains of the aluminium alloy matrix are of uniform size regardless of the location in the melt pool. (The top part of Figure 19 shows the melt pool structure - melt pool edges are white. This is due to poor orientation indexing, a result of locally higher dislocation density). More generally however, the microstructure of the materials in the invention may have multimodal grain size distributions and grain anisotropy. Typically, finer grains may cluster near the edges of semicircular regions (melt pools) and larger, sometimes elongated grains located in the centre of substantially semicircular regions. These semi-circular regions are formed during the melting phase of the AM process and are characteristic of the AM process. Examples 1 and 2 have a surface roughness lower than its equivalent made of the matrix alloy alone. The ceramic particles have the same properties as the ceramic particles of the powders. Figure 12 shows the optical density of items produced using the process in Example 2. It shows that a broad range of 3d energy density values (or VED) can result in very low porosity in the microstructure, indicating a broad processing window. The microstructure of Example 2 has many ceramic particles which are finer than the initial distribution due to the attrition during the milling process. It is to be expected when using the same method as used to make example 1, that the finer ceramic particles would not be present. 5 Figure 14 shows a comparison in machinability between Example 1 and 2. Samples of both materials were turned on a lathe with a surface speed of approximately 60 m s'1 using an uncoated cemented tungsten carbide tool. Example 2 wears down the tool slowly, indicating that Example 2 can be machined with cemented tungsten carbide tools, unlike 10 most of the state-of-the-art MMCs. This is due to the structure of the MAX phase compared to the TiB2. Figures 15-18 compare different mechanical and physical properties of Examples 1 and 2 and use matrix alloy 6061 and aluminium alloy AlSilO as benchmarks. Examples 1 and 2 15 show favourable strength, stiffness, density and ductility.

Claims

1. A method of manufacturing a metal matrix composite component, the method comprising:mixing a metal powder with a ceramic powder;binding the mixed metal powder and ceramic powder together to form an electrode; andusing the electrode in an electrode induction gas atomisation process to prepare a metal matrix composite powder, andmanufacturing a metal matrix composite component using the metal matrix composite powder in a metal additive manufacturing process.

2. The method of claim 1, wherein the metal additive manufacturing process is a powder-bed metal additive manufacturing process, preferably a laser powder-bed fusion additive manufacturing process, or wherein the metal additive manufacturing process is direct energy deposition printing.

3. The method of claim 1 or 2, wherein the ceramic powder is TiB2 or SiC or AIN ceramic powder.

4. The method of claim 1 or 2, wherein the ceramic powder is a MAX phase ceramic powder.

5. A method of preparing a metal matrix composite powder for additive manufacturing, the method comprising:mixing a metal powder with a MAX phase ceramic powder;binding the mixed metal powder and MAX phase ceramic powder together to form an electrode; andusing the electrode in an electrode induction gas atomisation process to prepare a metal matrix composite powder.

6. The method of claim 4 or 5, wherein the MAX phase of the MAX phase ceramic has a layered hexagonal structure, preferably a layered hexagonal crystallographicstructure with weak interlayer bonding and / or wherein the layered hexagonal crystallographic structure is a space group Pbs / mmc.

7. The method of any of claims 4-6, wherein the MAX phase of the MAX phase ceramic has a general formula Mn+iAXn, where M is one or more metals, A is one or more A-group element and X is C and / or N and / or B.

8. The method of any of claims 4-7, wherein M is one or more of: Ti, Sc, V, Cr, Zr, Nb, Hf, Mn, Mo and / or A is one or more of: Cd, Si, Sn, In, Ga, Tl, Al, Ga, Sn, Pb, Zn, Au.

9. The method of any of claims 4-8, wherein the MAX phase of the MAX phase ceramic is ThSiC?.

10. The method of any of claims 4-9, wherein the MAX phase of the MAX phase ceramic is A14SiC4 and AlsSiC?.

11. The method of any of the preceding claims, wherein the binding step comprises cold isostatic pressing.

12. The method of any of the preceding claims, wherein the binding step comprises sintering and / or the application of pressure.

13. The method of any of the preceding claims, further comprising sieving the metal matrix composite powder to achieve a predetermined particle size distribution.

14. The method of any of the preceding claims, wherein the volume of ceramic powder is at least 5% of the total volume of the sum of metal powder and ceramic powder, preferably wherein the volume of ceramic powder is at least 10% of the total volume of the sum of metal powder and ceramic powder, optionally wherein the volume of ceramic powder is 50% or less of the total volume of the sum of metal powder and ceramic powder.

15. The method of any of the preceding claims, wherein the metal powder is a powder of one or more of: nickel and nickel alloys, steels including stainless steels, titanium and titanium alloys, aluminium and aluminium alloys, cobalt and cobalt alloys, magnesium and 25magnesium alloys, copper and copper alloys, refractory metals and refractory metal alloys, platinum group metals and their alloys, vanadium metals and their alloys.

16. The method of any of the preceding claims, wherein the ceramic has a melting point in Kelvin at least 20% higher that the melting point in Kelvin of the metal of the metal powder.

17. The method of any of the preceding claims, wherein the metal powder is a powder of A16061.

18. The method of any of the preceding claims, wherein the metal matrix composite powder has the tenth percentile diameter (dio) between 10 pm and 25 pm and the ninetieth percentile diameter (d9o) between 40 pm and 90 pm.

19. The method of any of the preceding claims, wherein the metal powder has a dio of less than 15 pm and a dsio of 80 pm or less.

20. The method of any of the preceding claims, wherein the ceramic powder has a dio of 5 pm or less and a doo of 10 pm or less.

21. The method of any of the preceding claims, wherein the metal matrix composite powder has a circularity of 0.7 or more.

22. The method of any one of the preceding claims, wherein the mixing includes ball milling, preferably low energy ball milling.

23. The method of any one of the preceding claims, wherein the metal powder has a mean particle diameter which is larger than a mean particle diameter of the ceramic powder, preferably five times or more larger, more preferably ten times or more larger.

24. A metal matrix composite powder comprising:metal matrix composite particles, wherein each metal matrix composite particle comprises a metal matrix in which a plurality of MAX phase ceramic particles are embedded.

25. The metal matrix powder of claim 24, wherein the metal matrix composite powder has a circularity of 0.7 or more, preferably of 0.8 or greater.

26. The metal matrix composite powder of claim 24 or 25, wherein a volume fraction of ceramic in is in the range of from 5 to 50%.

27. The metal matrix composite powder of claim 24, wherein the metal matrix composite powder has the tenth percentile diameter (dio) between 10 pm and 25 pm and the ninetieth percentile diameter (d9o) between 40 pm and 90 pm.

28. The metal matrix composite powder of any of claims 24-27, wherein the ceramic particles have a dio of 5 pm or less and a d9o of 10 pm or less.

29. The metal matrix composite powder of any of claims 24-28, wherein the metal consists of 0-10% in sum of silicon, copper, iron, manganese, zinc, titanium, magnesium, lithium, zirconium, beryllium, scandium, sodium, cerium, yttrium, lithium, calcium, silver and the balance being aluminium and inevitable impurities.

30. The metal matrix composite powder of any of claims 24-29, wherein the MAX phase of the MAX phase ceramic has a layered hexagonal structure.

31. The metal matrix composite powder of any of claims 24-30, wherein the MAX phase of the MAX phase ceramic has a general formula Mn+iAXn, where M is one or more metals, A is one or more A-group element and X is C and / or N and / or B.

32. The metal matrix composite powder of any of claims 24-31, wherein M is one or more of: Ti, Sc, V, Cr, Zr, Nb, Hf, Mn, Mo and / or A is one or more of Cd, Si, Sn, In, Ga, Tl, Al, Ga, Sn, Pb, Zn, Au.

33. The metal matrix composite powder of any of claims 24-32, wherein the MAX phase of the MAX phase ceramic is TisSiC?.

34. The metal matrix composite powder of any of claims 24-33, wherein the MAX phase of the MAX phase ceramic is AhSiC4 and AlsSiC?.

35. An AM component characterised by a microstructure consisting of secondary phase particles of a ceramic powder distributed in a metal matrix.

36. The AM component of claim 35, wherein the ceramic particles exhibit limited clustering.

37. The AM component of claim 35 or 36, wherein the grain structure of the metal matrix is isotropic and monomodal or consists of finer grains near the edges of melt pool boundaries optionally with larger, optionally elongated, grains oriented normally to the melt pool boundaries.

38. The AM component of claim 35, 36 or 37, wherein the AM component is manufactured by the method of any of claims 1 to 23 or from the powder of any of claims 24-34.

39. A metal matrix composite manufactured by the method of any of claims 1 to 23.

Images