An overview of additively manufactured metal matrix composites: preparation, performance, and challenge

Metal matrix composites (MMCs) are frequently employed in various advanced industries due to their high modulus and strength, favorable wear and corrosion resistance, and other good properties at elevated temperatures. In recent decades, additive manufacturing (AM) technology has garnered attention as a potential way for fabricating MMCs. This article provides a comprehensive review of recent endeavors and progress in AM of MMCs, encompassing available AM technologies, types of reinforcements, feedstock preparation, synthesis principles during the AM process, typical AM-produced MMCs, strengthening mechanisms, challenges, and future interests. Compared to conventionally manufactured MMCs, AM-produced MMCs exhibit more uniformly distributed reinforcements and refined microstructure, resulting in comparable or even better mechanical properties. In addition, AM technology can produce bulk MMCs with significantly low porosity and fabricate geometrically complex MMC components and MMC lattice structures. As reviewed, many AM-produced MMCs, such as Al matrix composites, Ti matrix composites, nickel matrix composites, Fe matrix composites, etc, have been successfully produced. The types and contents of reinforcements strongly influence the properties of AM-produced MMCs, the choice of AM technology, and the applied processing parameters. In these MMCs, four primary strengthening mechanisms have been identified: Hall–Petch strengthening, dislocation strengthening, load transfer strengthening, and Orowan strengthening. AM technologies offer advantages that enhance the properties of MMCs when compared with traditional fabrication methods. Despite the advantages above, further challenges of AM-produced MMCs are still faced, such as new methods and new technologies for investigating AM-produced MMCs, the intrinsic nature of MMCs coupled with AM technologies, and challenges in the AM processes. Therefore, the article concludes by discussing the challenges and future interests of AM of MMCs.

Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Introduction
Metal matrix composites (MMCs) are the intimate combination of continuous metal matrix and reinforcing materials [1,2].MMCs encompass a diverse range of materials that can be classified based on their metallic matrix (e.g.Ti, Al, Mg, Cu, Fe, Ni, Co, etc) and reinforcing materials (e.g.particles, fibers, and whiskers) [3][4][5].Over the recent decades, MMCs have garnered global attention owing to their remarkable ability to exhibit improved toughness and good ductility within the metallic matrix, coupled with the high stiffness and strength conferred by the reinforcements.MMCs typically exhibit high modulus and strength, good wear resistance, and noteworthy high-temperature properties.Owing to these remarkable characteristics, MMCs have emerged as a good material choice and are employed in some industrial sectors [1,[6][7][8][9][10].For instance, the fan exit guide vanes utilized in the Pratt & Whitney 4000 series gas turbine engines, equipped on Boeing 777 commercial airliners, are constructed using discontinuously reinforced Al matrix composites [9].
Generally, conventional techniques for the fabrication of MMCs include stir casting, forging, diffusion bonding, infiltration, and powder metallurgy [11][12][13][14][15].The fabrication methods mentioned above invariably entail a multitude of processing steps.Furthermore, the attainment of the desired shape and dimensional accuracy of the part necessitates the involvement of machining, thereby contributing to the overall production cost of MMC components [16].Certain types of MMCs, such as Ti matrix composites, exhibit comparatively low thermal conductivity and a high degree of chemical reactivity.Therefore, undesired chemical reactions during the machining process negatively influence their machinability [17].The manufacture of MMCs using powder metallurgy is confronted with many challenges, primarily due to the interactions between the reinforcement and metal matrix, including weak interfacial bonding and non-uniform distribution of the reinforcements within the metal matrix [14].Surface and heat treatments are frequently used as post-processing techniques to enhance the properties of the as-fabricated MMCs [18,19].Consequently, endeavors have been undertaken to overcome these challenges and expand the scope of MMCs applications in recent years.Fortunately, additive manufacturing (AM) technologies have emerged as a promising alternative for producing MMCs [20].
AM technologies enable the layer-wise, computer-aided design (CAD) model-based, direct production of threedimensional components without any intermediate steps, originating in 1986 [21].These AM technologies alleviate the constraints imposed by conventional manufacturing methods, enabling the fabrication of components with more complex geometries.AM processes can potentially eliminate the requirement for extensive post-processing schedules, significantly reducing material waste and energy consumption associated with producing intended components [22][23][24][25].Additionally, the ease of generating CAD models facilitates the attainment of high-level designs that promote lean production practices.Under these benefits, AM technologies have emerged as a favored fabrication approach across various industries [26][27][28][29][30][31].
Besides their inherent characteristics, AM technologies have some merits in fabricating MMCs.There are two options for adding reinforcements during the AM process.The reinforcement can be pre-mixed with the metal matrix powder in various ways (such as ball-milling or mechanical mixing) to ensure the homogeneousness of the feedstock in the AM processes.Therefore, the reinforcements are not aggregated and have a homogeneous distribution in the AM-produced MMCs [32].Otherwise, the reinforcements can be added from powder feeders and mixed with the metal matrix powder during some AM processes.Subsequently, the mixed powders are melted by the heat source.In such AM processes, the proportional mass of the reinforcements and metal matrix powder can be precisely modulated by adjusting the feed rate of each respective powder feeder [33].Owing to this merit, the functional gradient MMCs can also be achieved [34].Moreover, some types of reinforcements are formed during the production of MMCs by in-situ reactions.Hence, precise control of in-situ reaction is the key to efficiently synthesizing desired reinforcements.The energy input can be controlled during the AM process by optimizing the processing parameters [35].As such, the undesired chemical reactions between the additive and matrix can be partially avoided.
Various AM technologies have successfully fabricated numerous fully dense MMCs, including but not limited to Ti-, Al, Fe-, Ni-, and Co-based MMCs, displaying promising performances [6,18,19,[36][37][38].Consequently, AMproduced MMCs have gained considerable attention in recent years.This article comprehensively overviews recent developments and research efforts in AM-produced MMCs.The primary focus is on the principles of AM technologies used for MMC production, feedstocks, and reinforcements, as well as the properties and applications of various MMCs.Additionally, the strengthening mechanisms of MMCs are discussed.Based on current research findings, this review also discusses the opportunities, challenges, and prospects of AMproduced MMCs.

Available AM technologies for producing MMCs
In recent decades, numerous AM technologies have emerged and been used for fabricating MMCs [39].In order to obtain high-performance MMCs, it is crucial to achieve homogeneous reinforcement distribution [6].However, due to their different processing principles, not all AM technologies are appropriate for producing MMCs.The primary feedstocks used in AM technologies include particles, fibers, whiskers, wires, and sheets.When employing wires or sheets as feedstocks, it is imperative to introduce potential reinforcements (particles, fibers, and whiskers) before manufacturing the wire or sheet.
Plastic deformation is typically involved during the production of wire or sheet feedstocks, which may result in uneven reinforcement distribution within the material [40].Deng et al [41] addressed this issue by preparing in-situ TiB 2 particles that reinforced Al-7Si-Cu-Mg composite ingots, which were extruded and drawn into wire feedstocks.The resulting wire feedstocks still exhibited suboptimal reinforcement distribution.Using feedstocks with non-uniform reinforcement distribution is not advisable for fabricating highperformance MMCs.Hence, powder-based AM technologies remain the primary choice for producing MMCs.In the subsequent sections, several powder-based AM technologies will be introduced, along with their applications in the fabrication of MMCs.
The feedstock in the PBF processes can be composite powder or mixture.Therefore, in many cases, one can conclude that the reinforcements are integrated before the PBF process [68].Of course, there are also exceptions.The highenergy heat sources of PBF technologies facilitate in-situ reactions, enabling the formation of reinforcements by the in-situ reaction between the metal matrix powder and additives [36,69].It has also been reported that the reinforcements can be produced via reactive gases in the processing chamber, reacting with metal matrix particles [70].PBF has a good accuracy in production but a low production rate.Therefore, PBF is particularly advantageous for the parts with complex geometries.Moreover, PBF has a high material utilization, allowing for the recycling of non-sintered feedstock [71].Given the substantial cost of composite feedstock, such recycling practices are critical in minimizing the overall production expenditure.Apart from PBF technologies, other AM technologies are also accessible for fabricating MMCs.

Direct energy deposition (DED)
Unlike PBF technologies, DED technology uses blown powder feedstock with a scanning high-energy source to produce components based on a pre-designed CAD model [72][73][74][75].This technology is often regarded as a developed laser cladding technology in the literature [76][77][78].During the DED process, the deposited feedstock is melted in an inert gas-protected environment, creating an active melt pool.Multiple nozzles can be employed to deposit different materials simultaneously [72].Therefore, DED can process various materials, including alloys, ceramics, and composites [72,73,[79][80][81].However, DED-produced parts often exhibit poor surface finish and low build accuracy, with partially melted powder commonly found on the components.The attached partially melted powder can increase surface roughness and decrease fatigue performance [82][83][84].Therefore, post-processing treatments (e.g.hot isostatic pressing (HIP), heat treatment, and machining) are typically required to enhance the performance of DED-fabricated components [85][86][87].
DED technology exhibits several advantages in the production of MMCs.Fundamental to the DED technology is the capacity for simultaneous deposition of different materials using multiple nozzles [72].During this process, reinforcements can be introduced alongside the metal matrix powder, allowing for the precise regulation of different material ratios through the controlled adjustment of the flow rate of each powder feeder [88].Multiple reinforcements can especially be added during the DED process, and therefore, the MMCs with multiple reinforcements can be produced.The strategic control of powder feed rates within the DED process makes the creation of functionally graded MMCs attainable [34].Typically, the energy input density in DED exceeds that of PBF, resulting in a greater production rate for DED than PBF [89].PBF and DED technologies utilize a high-energy source to melt materials during the process.By contrast, alternative AM technologies do not utilize a heat source but employ a liquid binding agent and bond materials to produce MMC components.

Binder jetting (BJ) printing
The fabrication method known as BJ printing necessitates using a powder bed, like that employed in PBF techniques.BJ printing entails the application of a liquid binding agent onto the powder bed to produce a two-dimensional pattern [90][91][92].The layer-by-layer deposition of the binding agent leads to the production of a 3D component.The BJ printing could employ any powder-based feedstocks (such as metals [93], ceramics [94], polymers [95], and composites [96]) in combination with a binder to produce components.Furthermore, BJ printing can be performed in the air, making it a cost-effective and highspeed AM process compared to other methods [97].However, owing to the binder, post-processing processes (e.g.sintering, de-binding, infiltration, and powder removal) are required to remove the binder from the final products [98][99][100][101].Removing the binder leaves pores produces relatively high porosities in the BJ-fabricated components.Hence, HIP is commonly used to minimize porosity [102].Even so, full density is still unachievable with BJ-printed components, resulting in lower mechanical properties than their numerical models, even if well-designed [103].
In the BJ process, the absence of a heat source precludes the possibility of in-situ reactions during fabrication, necessitating the pre-integration of reinforcements within the feedstock in most instances [104].Owing to such a feature of BJ, the prealloyed powder is better to produce MMCs with good performances.Nevertheless, in-situ reactions can be facilitated during post-processing stages, such as sintering, which removes the binder.Do et al [101] demonstrated that the employment of Inconel 625 alloy with a carbon-containing binder in the BJ process results in the formation of carbides through reactions between carbon and elements in the alloy during the sintering.Moreover, the binder should be removed, which produces pores in the final products, resulting in decreased mechanical properties [103].BJ may be considered analogous to PBF from certain perspectives, with the former bonding materials using a binder and the latter employing a high-energy heat source for bonding.Both PBF and BJ exhibit impressive competencies in manufacturing parts with complex geometries, albeit their production rates are comparatively moderate.

Spraying technology
Compared to PBF and BJ technologies, spraying technologies have high production rates for manufacturing parts, encompassing metals and MMCs.Spraying technologies for manufacturing MMCs can be categorized into thermal and cold spraying methods [105][106][107].Both techniques employ composite powder as feedstock to fabricate MMC components.Thermal spraying methods use the flame, electric arc, or plasma arc as heat sources and spray the powder in a semi-molten state in high-speed gases [108][109][110].The temperature of the high-speed gases ranges from 1500 • C to 15 000 • C, depending on the heat sources used [108,110].Cold spraying employs a high-pressure supersonic gas jet to propel powder particles at or surpass a critical velocity range of 500-1200 m•s −1 to fabricate components [111].Cold spraying can occur at room or low pre-heated temperature (typically below 600 • C) [112].Both thermal and cold spraying processes involve impact and deformation, and the produced components often contain pores and lamellar boundaries that negatively affect their performances [113,114].As a result, post-treatments (e.g.remelting and heat treatment) are sometimes required [109,113].In addition, the kinetic energy of in-flight particles during cold spraying is significantly higher.Such a phenomenon results in strong particlesubstrate and particle-particle interactions, effectively rupturing surface oxides and promoting clean interfaces [115][116][117].Consequently, the components manufactured using cold spraying exhibit fewer flaws than those produced via thermal spraying.Nevertheless, to achieve MMCs with optimal performance, the choice of spraying technology should be fully evaluated based on the characteristics of the feedstock.
Both pre-alloyed and mixed powders are viable feedstocks for fabricating MMCs through thermal or cold spraying technologies [115,118].Applying a heat source enables in-situ reactions to occur during the process, as shown in [119].Specifically, during spraying, additives may undergo reactions that result in the formation of reinforcements within the produced components.Additionally, reactive gases have been reported to interact with materials during thermal spraying, forming reinforcements [120].Nevertheless, using reactive gases can adversely affect the anode nozzle of thermal spraying equipment.
Spraying AM differs from other AM techniques, particularly regarding product accuracy.However, it is highly efficient in producing coatings, making it a more appropriate choice for such applications.While CAD can also be utilized in the context of spraying AM, the achieved accuracies of the products are typically lower compared to those produced by other AM technologies such as PBF and DED.In PBF and DED processes, molten pools are consistently present on the surfaces of the produced parts throughout the manufacturing process [121].In contrast, there are no molten pools during spraying AM.The feedstock is melted or semi-melted at the nozzle of the spraying AM equipment, and the resultant melted or semi-melted in-flight particles are sprayed onto the substrate following a pre-designed route [122].Consequently, the designed deposition of in-flight particles cannot be guaranteed.
Selecting the appropriate AM technology is paramount to achieving good production of MMCs since each technology presents its distinct set of advantages and disadvantages.Table 1 outlines several AM technologies that produce MMCs and the corresponding processing mechanisms.

Feedstock
Indeed, despite utilizing powder, wire, and sheet as feedstocks in the AM process for fabricating metals and alloys, it is noteworthy that, in the context of MMCs, the predominant feedstocks employed are generally powders [6,38].Various composite powders have been reported for different AM processes, including gas-atomized powder, mechanically mixed powder, ball-milled powder, and satellited powder [18,[123][124][125][126].Four commonly used typical powders are shown in figure 1.The gas-atomized powder is produced by incorporating reinforcements within the metallic powder (figure 1(a)) [18].The composite ingots are melted and gas-atomized to produce the High porosity, poor surface finish, very low accuracy; parts have limited shapes; the even-distributed reinforcements are very difficult to achieve using the mechanically mixed feedstock composite powder.Gas-atomized powder generally exhibits good flowability and sphericity but is expensive [37,127].Mechanically mixed powder is generated by rotating a mechanical mixer for several hours, resulting in unevenly distributed reinforcements in the MMCs due to the non-attached or detached particles (figure 1(b)) [123,[128][129][130][131]. Mechanically mixed powder often shows poor wettability, resulting in the instability of the melt pool during the AM process [132].
The ball-milled powder subjected to ball milling is obtained through hardened milling balls, which are processed via either a planetary ball mill or a ball-tube mill.The process is strictly controlled to prevent deformation or non-attachment of reinforcements in the final composite powder (figure 1(c)) [133,134].It has been reported that gases can become trapped inside the particles during the powder atomization process, forming pores [135].Ball milling is effective at sealing these pores [20].However, ball milling can adversely affect the surface finish of MMC components.AlMangour et al [136] also found that ball-milled powder exhibits more extensive and severe cracking than directly mixed powder.However, the directly mixed powder demonstrates the presence of deep grooves and pores.Satellited powder, also known as wet granulation, is produced by adding a small amount of binder, such as polyvinyl alcohol, to the blender of metal particles and reinforcements, followed by gyroscopic mixing to obtain the desired powder (figure 1(d)) [125,137].Although satellited powder exhibits a higher level of reinforcement attachment, the poor adhesion of the binder can still result in unsatisfactory adhesion of reinforcements.Mechanically mixed powder, ball-milled powder, and satellited powder have inferior flowability and sphericity compared to gas-atomized powder.Therefore, the gas-atomized composite powder is a good choice for fabricating high-performance MMCs.However, the production cost of gas-atomized powders is significantly higher than those of other composite powders, necessitating considering the performance-cost tradeoff.Table 2 lists the advantages and disadvantages of these types of composite powders.These composite powders employ distinct methods to integrate the metal matrix powder and reinforcements.The specific approaches vary in their techniques for achieving a homogeneous mixture, ensuring proper distribution of reinforcements within the metal matrix.The subsequent section will delve into the unique characteristics of these reinforcements, exploring their role in enhancing the properties of the resulting MMCs.

Reinforcements
Incorporating reinforcements is an important factor that influences the fabrication and properties of MMCs, as it serves to refine their microstructures and inhibit the formation and propagation of cracks [129,[138][139][140].For MMCs to exhibit high performances, the reinforcements should possess high thermal stability and hardness and be compatible with the metal matrix.Successful MMC design of the AM procedure requires the consideration of numerous factors, including the chemical composition of reinforcements, the natural oxide film, the flowability of feedstock, and processing parameters during the AM process.MMCs produced using AM and conventional methods can be categorized based on two methods of introducing reinforcements: ex-situ and in-situ [38,141].
The ex-situ method involves adding the reinforcements to the metal matrix before the AM process.In contrast, the in-situ method produces reinforcements during the AM process via the in-situ reaction between the metal matrix and additives.
Compared with the ex-situ method, the in-situ method can promote the nucleation and growth of reinforcements and prevent the formation of interfacial products [128,142].In-situ reinforcements are also more uniformly distributed.The in-situ reinforcements can be exemplified by the production of TiB/Ti composites, where TiB is generated by the in-situ reaction between TiB 2 and Ti matrix [143].The mixture powder comprises TiB 2 particles surrounding the Ti particles (figure 2(a)).During the process, the in-situ reaction results in the generation of TiB (figure 2(b)).Although some unreacted TiB 2 particles remain, the resulting TiB particles are evenly dispersed.Notably, the processing parameters can influence the in-situ reaction.
Moreover, the in-situ method confers a notable benefit in the fabrication of MMCs owing to the strong interfacial adhesion engendered between the reinforcements and the metal matrix as an outcome of the in-situ reaction, as reported in [144,145].Consequently, MMCs fabricated by the in-situ method generally exhibit superior properties to those fabricated by the ex-situ method [144,145].However, the AM process can cause instability in the melt pool of materials during the in-situ reaction, which generates additional heat and causes excessive evaporation and powder splash [36,146].Therefore, when designing and selecting a material system for in-situ composites in the AM process, several critical factors, such as the metallurgical process, thermal history of materials, and compositions of additives and matrix, should be carefully considered and designed.
The selection of reinforcements for MMCs should consider various factors, including their dimensions, morphology, coefficient of thermal expansion (CTE), elastic modulus (EM), crystallographic structure, wettability, and in-situ reactivity with the metal matrix.For the ex-situ method, the dimension and morphology of the reinforcements play a crucial role in their distribution in the blend, thereby affecting the performances of the produced composites [147].Saba et al [147] investigated the effect of diamond powder size on the Ti matrix.They found that adding diamond powder with an average size of 5 nm could significantly improve the strength and ductility of in-situ composites by interacting with dislocations and inhibiting grain growth.Moreover, the mismatch of the coefficients of thermal expansion and elastic moduli between the reinforcement and matrix produces geometrically necessary dislocations (GNDs), which can improve the strength of composites [148,149].However, the mismatch of the coefficients of thermal expansion between reinforcement and matrix can result in residual stress at their interfaces, which becomes the site for crack nucleation [150].The crystallographic structure of the reinforcement determines the lattice mismatch with the metal matrix, which further facilitates the heterogeneous nucleation of melt during solidification [151].For instance, Xi et al [151] found that TiB 2 ceramic particles with low lattice match with the Al matrix could eliminate the pronounced crystallographic texture of LPBF-produced TiB 2 /Al-12Si composite, resulting in a homogeneous microstructure with refined equiaxed grains with random orientations.A favorable wettability of the molten metal matrix can enhance the interfacial strength between the metal matrix and reinforcement [152].
The in-situ reaction of the additives and metal matrix ultimately determines the final reinforcements in the composite [153].Table 3 summarizes the characteristics of widely used reinforcements or additives, while table 4 describes the characteristics of the frequently employed metal matrix.New MMCs can be designed based on the characteristics of the additives and metal matrix.To successfully produce MMCs using AM technologies, a judicious selection of an appropriate metal matrix combined with suitable additives is crucial.For AM technology development, numerous instances of successful printing of MMCs have been reported, showcasing the feasibility and potential of this approach in fabricating advanced composite materials.

Matrix materials
Metallic powders serve as the primary feedstocks in the AM process.The consistency of the printing, when viewed through the powder characteristics, such as particle size distribution, flowability, and sphericity, plays pivotal roles in influencing the spreadability and recoatability during the AM process [71].These factors, in turn, can have consequential effects on the densifications, microstructures, and performances of the parts.Apart from pre-alloyed composite powder, other types of composite powders, including mechanically mixed powder, ball-milled powder, or satellited powder, are all derived from the powder made of matrix material.Therefore, the quality of the metal matrix powder, such as particle size distribution, sphericity, flowability, and chemical composition, is vital for preparing composite feedstocks.Particle size distribution delineates the proportionality of particles of disparate sizes in the powder, influencing the packing efficiency, particle melting, and flowability of particles [172][173][174].AM technologies impose distinct requisites on powder particle size parameters [175].Sphericity indicates the degree to which the particle morphology approximates a spherical geometry; the sphere, optimal due to its minimized surface-to-volume ratio, facilitates reduced surface friction traction and enhances flowability, spreading behavior, and packing density [176].Flowability is characterized by powder mobility and significantly influences homogenous powder deposition, resulting in consistent layer thickness, packing density, and uniform energy assimilation, which, in turn, augments the density and functional attributes of the final products [177].Chemical composition is paramount due to its capacity to precipitate pronounced changes in product density, phase distribution, and performance with even minuscule alterations [178].Therefore, the characteristics of composite feedstocks are distinctly influenced by the attributes above of the metal matrix powder.
For illustrative purposes, utilizing the metal matrix powder with a narrow particle size distribution in feedstock preparation is undeniably different from the powder characterized by a broad particle size distribution in the performances of the final products [179].
As reviewed, many metals and alloys can be used as the metal matrix to produce MMCs, such as Al, Ti, Ni Fe, Co, and their alloys.The widely used methods for producing metal matrix powder are gas atomization [180], electrode induction melting gas atomization [181], vacuum induction melting gas atomization [182], plasma atomization [183], water atomization [184], centrifugal atomization [185], and plasma rotating electrode process [186].Irrespective of the powder production methods, there is a widely acknowledged consensus that the cost of powder employed in the AM process is high [71].Consequently, the efficient reduction of powder costs emerges as a pivotal strategy for enhancing the overall cost-effectiveness of the fabrication process.An approach to mitigate powder costs involves reusing the remaining powder after the AM process.Nonetheless, it is imperative to recognize that the recycled powder may undergo change in terms of physical and mechanical properties, encompassing aspects such as chemical composition, particle size distribution, and flowability [187,188].This change is attributed to the interactions between the heat resource and powder during the AM process.Therefore, caution should be exercised when using recycled powder to prepare composite feedstock, as it may yield different results than composite feedstocks prepared using virgin powder.
Researchers have employed virgin powder to prepare the composite feedstock in most published works.Almost no works regarding recycled powder are mentioned.However, recycled powders are used more frequently for cost consideration, even if commercialized [189,190].Therefore, it is still worthy of understanding the difference between virgin powder and recycled powder [191].Huang and Yan [71] concluded the characteristics of the reused powder in terms of particle size distribution, sphericity, and flowability.The primary factors contributing to the altered properties observed in the powder upon recycling are the genesis of spatters and a reduction in the proportion of fine particles [192,193].Therefore, the reused powders have a right-shifted particle size distribution, lower sphericity, and flowability compared to the virgin powder [71].In order to mitigate the influence of the spatters, the reused powders are frequently subjected to sieving or exposure to airflow, effectively eliminating agglomerated spatters.In such a situation, the reused powders would have a left-shifted particle size distribution and recovered sphericity and flowability [192,194,195].It follows that recycled powders, despite treatment, retain distinctive attributes when contrasted with virgin powder.The properties of the utilized powder should be considered when preparing composite feedstock.Detailed characterization of the metal matrix powder is advocated for augmented quality control of the composite feedstock.

Aluminum matrix composites (AMCs)
AMCs exhibit properties derived from Al alloys and reinforcements, enabling AMCs to demonstrate high specific strength and stiffness, low CTE, and good wear resistance [68,[196][197][198].As such, AMCs have become commonly used lightweight materials in various industrial sectors.Conventional AMCs can be produced via melt processing or powder metallurgy [199][200][201][202][203][204][205][206].Despite the promising properties of conventional AMCs, they have been observed to exhibit certain limitations, including irregularly distributed reinforcements, inadequate interfacial bonding between the matrix and reinforcement, and high levels of porosity [207][208][209].Recently, AM technologies have been employed to fabricate particle-reinforced AMCs, which can overcome these issues.Compared to conventional AMCs, AM-produced AMCs exhibit a relatively even distribution of reinforcements, low porosity, and a clear matrix-reinforcement interface.The most commonly used reinforcements in AMCs include SiC, TiC, AlN, BN, Si 3 N 4 , Al 2 O 3 , ZrO 2 , and TiO 2 [38,124,210].
Carbides are commonly used as reinforcements in AMCs.Gu et al [211] reported a comparative investigation on the tensile properties of LPBF-produced TiC/AlSi10Mg composite and LPBF-produced AlSi10Mg alloy.They found that the tensile strength of LPBF-produced TiC/AlSi10Mg composites is 52 MPa greater than that of their counterparts.Raj Mohan et al [212] prepared NbC/AlSi10Mg composites with different NbC particles by LPBF.They demonstrated that the LPBFproduced NbC/AlSi10Mg composites exhibit better mechanical properties, including tensile strength, hardness, and wear resistance, compared with the unreinforced LPBF-produced AlSi10Mg alloy.The addition of NbC does not decrease the grain size of the composites, and dispersion strengthening is identified as the primary mechanism responsible for strengthening LPBF-produced NbC/AlSi10Mg composites.An interesting phenomenon was found by Chang et al [213].They used varied sizes of SiC particles (d 50 = 5 µm, 15 µm, and 50 µm, respectively) to fabricate AMCs by mixing AlSi10Mg powder and employing the LPBF process.As revealed in figure 3(a), the formation of in-situ reaction products (Al 4 SiC 4 ) is discernible only in the composites containing fine and mediumsized SiC particles.Despite all feedstocks having an identical SiC particle content (20 wt%) and applying identical LPBF process parameters, the composite with large SiC particles exhibits the a reduced relative density (figure 3(b)).Therefore, this particular composite demonstrates the diminished hardness coupled with an escalated wear rate (figures 3(c) and (d)).The fine particles (SiC) in the melt provide more interfaces and facilitate the in-situ reaction [214].Hence, as the initial particle size diminishes, in-situ reactions intensify, resulting in greater reinforcement formations.Due to the notably brief period of laser interaction, the large SiC particles are still almost unmelted.Such a situation interrupts and causes discontinuities within the laser-scanned molten tracks [214,215].Hence, at this stage, the viscosity and surface tension within the molten pool are comparatively elevated, leading to a markedly chaotic melt flow.Such chaotic melt flow results in the non-uniform dispersal of both mass and thermal energy.The resultant destabilization of the melt pool, alongside the pronounced balling effect, contributes to the emergence of large pores adjacent to the SiC reinforcements.Hence, the careful selection of reinforcement particle size is pivotal for the AM process.
Besides the direct addition of reinforcements, the in-situ reaction is also reported in the fabrication of AM-produced AMCs.Yi et al [69] used CP-Ti, B 4 C, and AlSi10Mg particles (B 4 C-Ti content from 0.7 wt% to 17.2 wt%) as initial materials during the LPBF process.A notable in-situ reaction occurs between CP-Ti and B 4 C, resulting in the generation of TiB 2 and TiC within the produced AMCs.Compared to unreinforced LPBF-produced AlSi10Mg alloy, the tensile strength of the composite with the incorporation of 0.7 wt% CP-Ti and B 4 C experienced an increase.With increased content of CP-Ti and B 4 C, the tensile strength of the composites decreases.Incorporating additional additives correlates with an augmented exothermic response from in-situ reactions.This phenomenon contributes to the remelting of the interface between melt pools and diminishes the viscosity of the liquid phase as the operative temperature within the molten pool rises [20].The consequence of these conditions is the emergence of an enlarged molten pool, with an augmented liquid phase temperature and an extended residency time within the pool.It is beneficial for the even dispersion of AlSi10Mg melt and leads to superior material densification [216].However, an elevation in operative temperature leads to an increment in the size and quantity of reinforcements.Crack initiation and propagation are readily facilitated at the interface between the soft aluminum matrix and the hard reinforcements under tensile stress, resulting in brittle failure and a consequent diminution of tensile strength.
By contrast, Xi et al [217] employed LPBF to fabricate (ZrC + TiC)/Al composites using mixed Al, ZrC, and TiC powder.An in-situ reaction between ZrC and TiC is initiated in the LPBF process to generate (Ti, Zr)C phase, which facilitates the formation of strong bonding with the Al matrix [218].Figures 4(a)-(e) confirm the even dispersion of TiC-rich and ZrC-rich particles in the (ZrC + TiC)/Al composites produced by LPBF.Transmission electron microscopy (TEM) investigations of the TiC-rich and ZrC-rich particles reveal clear interfaces, as shown in figures 4(f)-(j).In the domain of the interfaces between the Al matrix and reinforcements, the Ti, Zr, and Al elements change gradually, as illustrated in figures 4(h) and (k).The mechanical tests conducted indicate that the produced composites exhibit enhanced strength and wear resistance.Well-bonded interfaces between the matrix and reinforcements resulting from in-situ reactions play a critical role in determining the performance of the composite.However, the in-situ reactions are often exothermic.Such heat input produces an extra factor influencing the overall AM process [69].Therefore, the design of MMCs should comprehensively consider all potential factors.
Other reinforcements, including multiple types, have been employed in the fabrication of AMCs by AM technologies.For example, Wang et al [124] used ball milling to prepare TiB 2 /AlSi10Mg feedstock, which was subsequently used in the LPBF process to produce TiB 2 /AlSi10Mg composites.The porosity levels of the produced composites are highly dependent on the milling dose, which indicates the milling progress and is regulated as the ratio of the impact energy transferred to the powder mixture from the mass of milled powder [219].The resultant TiB 2 /AlSi10Mg composite exhibits a density of 99% at the milling dose of 275 W s g −1 , improving tensile strength (450 MPa) and ductility (7.2%).Mair et al [220] used gas-atomized Al-Cu-Ag-Mg-Ti-TiB 2 powder in the LPBF process to fabricate crack-free Al-Cu-Ag-Mg-Ti-TiB 2 (A205) samples with a relative density of (99.5 ± 0.1)%.TiB 2 reinforcements are homogeneously dispersed in the sample, and their sizes ranged from 0.1 µm to 1.3 µm (figures 5(a)-(c)).Al 2 Cu phase is distributed at the grain boundaries, and TiB 2 particles are embedded in the grains (figures 5(d)-(f)).The Al 2 Cu and TiB 2 particles exhibit distinct and coherent interfaces with the Al matrix (figures 5(g)-(n)).More examples of the properties of AM-produced AMCs with different reinforcements are shown in table 5 in the following.
Besides the well-known dispersion strengthening introduced by reinforcements, the disparity in thermal expansion coefficients between the metal matrix and the reinforcements generates GNDs within MMCs.This phenomenon results from the differential expansion and contraction rates between the matrix and the reinforcements during thermal cycling [221].The introduction of GNDs is a crucial factor influencing the microstructural evolution and mechanical properties of the composite, highlighting the intricate interplay between the metal matrix and the reinforcements in MMCs.Therefore, the produced GNDs impede each other, giving rise to the dislocation strengthening of the metal matrix.The dislocation strengthening could be expressed using the following equation [222]: where G is the shear modulus of the matrix, b is Burger's vector, M is the Taylor factor, B is a constant coefficient, and ρ is the dislocation density.ρ is respectively derived from the mismatch in the CTE and EM, which could be given by [6,223]: where ρ CTE and ρ EM are the dislocation densities resulting from the mismatch in the CTE and EM, respectively, of the matrix and reinforcement, ∆T is the difference in the processing temperature and ambient temperature, ∆α is the difference in the CTE between the matrix and reinforcement.
For ∆σ M , the calculations from ρ CTE and ρ EM are required, respectively.In [222]  strengthening mechanism plays an important role in strengthening the MMCs, thermal stresses, engendered by the differential CTE between the metal matrix and the reinforcements, may induce cracking and interlayer de-bonding.This tendency is exacerbated by residual stresses being more pronounced in the direction orthogonal to the scanning direction than in the parallel direction [226].Therefore, it is essential to meticulously regulate processing parameters to mitigate residual stress levels.Furthermore, AMC coatings are commonly produced using thermal spraying technologies.Xie et al [18] developed a novel 7075Al coating reinforced with nano-TiB 2 particles using cold spray AM with atomized TiB 2 /7075Al composite powder.The even distribution of TiB 2 nanoparticles in the composite powder improves the mechanical properties of the sprayed TiB 2 /7075Al composite coatings, including hardness, elongation, and tensile strength, compared to those of sprayed 7075Al coatings using the same spraying parameters.Similarly, Zhang et al [227] prepared It is imperative to note that inadequate inter-particle bonding inevitably leads to premature failure of Al composite coatings in thermal spray AM [229][230][231].As a result, post-friction stir processing is frequently employed after preparing coatings to further enhance their properties.Xie et al [229]   work-hardening capacity of TiB 2 /AlSi10Mg is also slightly enhanced (figure 8(f)).The post-treated parts demonstrate an excellent balance of strength and ductility.Similar investigations are also conducted in [230,231].Post-friction stir processing significantly decreases the porosities of cold-sprayed Al 2 O 3 /Al composite coatings, and their microstructures are homogenized [230,231].
Al alloys are widely used in various industrial applications, ranking second only to steel in terms of usage.Therefore, AMCs have been significantly developed to overcome the shortage of Al alloys.In addition to the examples discussed in this article, numerous investigations on the production of AMCs by AM have been reported.Table 5 gives an overview of the mechanical properties of AM-produced AMCs and reveals certain consistent trends.Analogous to the AMproduced TMCs, the strength of AM-produced AMCs also exhibits an ascending trend with higher reinforcement content [69,[232][233][234][235][236][237][238][239].Notably, AMCs fabricated through LPBF exhibit superior strength compared to their counterparts produced via DED [240][241][242].It is crucial to emphasize the substantial influence of processing parameters on the properties of the generated AMCs; variations in processing parameters may lead to significantly distinct material properties [211,[243][244][245].Moreover, the feedstock preparation also significantly influences the final performances of the produced composites [236-238, 246, 247].The pre-alloyed powder as feedstock to produce composites often performs better than composites produced using mechanically mixed or ball-milled powder [236][237][238].As mentioned above, the well-boned interfaces between the metal matrix and reinforcements can decrease the nucleation sites of cracks.Therefore, obtaining the well-boned interfaces between the metal matrix and reinforcements may be considered from the feedstock preparation stage.
TiC exhibits comparable density, Poisson's ratio, and thermal expansion coefficient to Ti alloys.Nevertheless, TiC outperforms Ti alloys significantly in modulus and strength [279,283].Furthermore, neither conventional nor AM methods reveal any visible reactions between TiC and Ti matrix, thereby establishing the wide utilization of TiC in TMCs [280,284].Gu et al [285] employed ball-milled TiC/Ti powder with different weight percentages of TiC (7.5 wt%, 12.5 wt%, 17.5 wt%, and 22.5 wt%) to fabricate TiC/Ti composites using LPBF.The relative densities of TiC/Ti composites decrease with increased TiC content.Notably, TiC/Ti composites with 12.5 wt% exhibit optimal comprehensive properties, including increased microhardness (577 HV 0.2 ) and a low friction coefficient (0.19).Gu et al [132] also prepared TiC/Ti composites using LPBF using ball-milled and mechanically mixed TiC/Ti powder, respectively.Using identical processing parameters, the TiC/Ti components fabricated from ball-milled powder exhibited higher mechanical properties and relative densities than those produced from mechanically mixed powder.This phenomenon is attributed to the superior densification of ball-milled TiC/Ti powder during the LPBF process.Ma et al [286] fabricated TiC/Ti composites with TiC content ranging from 10 wt% to 50 wt% using ball-milled TiC/Ti powder and the DED process.The TiC particle and α-Ti matrix show a semi-coherent interface with the orientation relationships of ] α-Ti // [110] TiC and TiC (figure 9).As a result, the nucleation of Ti and a well-bonded particlematrix interface is facilitated.Even though the addition of TiC enhances the hardness and wear resistance of TMCs, it leads to an inverse trend in their tensile strength and elongation.Recently, Wei et al [287] developed an in-situ laser AM technology to produce agglomeration-free TMCs with nano-scale TiC through gas-liquid reactions using CH 4 .The nanoscale TiC dispersions possess a greater specific surface area, providing a significantly enlarged interface area with the matrix compared to conventional composites.A characteristic that is conducive to an enhanced pinning effect [288].The mass fraction of TiC, as well as the yield strength of the composite, increases with CH 4 used.Nevertheless, as the CH 4 concentration escalates to 23 vol%, there is a noted decline in both the ultimate compressive strength and the plasticity of the TiC/Ti6Al4V composite.This trend underscores the premise that excessive incorporation of brittle TiC precipitates can amplify the brittleness of the composites.
Unlike TiC, TiB can be produced by in-situ reaction during the preparation of composites via the following reaction [289,290]: The in-situ production of TiB during the fabrication of composites facilitates a strong metallurgical bonding with the Ti matrix, and the resulting TiB is typically fine and uniformly dispersed throughout the matrix [291].As a result, the interest in investigating AM-produced TMCs with in-situ TiB reinforcements increases.Hu et al [289,292] employed a mixture of 1.6 wt% B powder and 98.4 wt% pure Ti powder to manufacture TiB/Ti composites using DED and obtained a TiB/Ti composite with the TiB content of 8.5 vol%.TiB particles form a fully quasi-continuous network, resulting in improved mechanical properties than those without a 3D quasi-continuous network.
Similarly, Cai et al [290] employed TiB 2 as the additive in the LPBF process to produce TiB/Ti-6Al-4V composites, where TiB 2 is transformed into needle-like TiB.The presence of TiB results in enhanced hardness and EM of the TiB/Ti-6Al-4V composites with an increased TiB fraction.Pan et al [293] used TiB 2 as the additive in the EBPBF process to produce TiB/Ti-6Al-4V composites.The in-situ reaction between TiB 2 and the matrix generates a quasi-continuous network of nanosized TiB whiskers within the composite microstructure.Such a microstructural configuration contributes to the enhanced tensile strength and plasticity of EBPBF-produced TiB/Ti-6Al-4V composites owing to the dispersion strengthening (figures 10(a) and (b)).The EBPBFproduced TiB/Ti-6Al-4V composite exhibits TiB whiskers arrayed along the grain boundaries in the metal matrix.This arrangement induces an altered fracture mode, ultimately leading to a postponement in the advancement of cracks (figure 10(c)).Cracks are formed when too many dislocations pile up and get tangled at the boundaries of the grains in the EBPBF-produced TiB/Ti-6Al-4V composite.Upon the application of stress, the microsized TiB whiskers in the forged counterpart fractured into small fragments, resulting in the production of microvoids.Therefore, the microsized TiB whiskers are the nucleation sites for cracks and are a major cause of the material breaking.The enhanced tensile strength and ductility exhibited by the EBPBF-produced TiB/Ti-6Al-4V composites can thus be attributed to the  rapid cooling rate inherent to the EBPBF procedure, which effectively refines the TiB whiskers synthesized via in-situ reactions.
Recently, a novel method for fabricating TMCs using reactive atmosphere during the LPBF process was developed [70,[294][295][296].For the conventional LPBF process, the reinforcements are prepared in the feedstock in advance, regardless of whether they are pre-alloyed powder or mixed powder.Xiao et al [295,296] produced TiN/Ti and TiN/Ti6Al4V composites with spatial heterostructures by LPBF under a nitrogen-containing atmosphere.The produced composites show increased mechanical strength without a significant decrease in ductility.The TiN-Ti heterolayer net-like structure stands out by attaining a high ultimate tensile strength of approximately 1.0 GPa and an elongation of 27% [70].This finding showcases a superior strength-ductility combination compared to intrinsic pure Ti, uniform TiN composites, and traditional layered structure Ti-based composites.The interaction between the TiN-Ti melting track forms net-like heterostructures with a higher density of soft-hard interfaces, resulting in excellent plasticity and strain hardening.Through insitu synthesis, the development of spatially heterostructured Ti composites notably improves the effective control of both the strength and plasticity of pure Ti in a balanced manner.Similar works regarding the AM-produced TMCs using reactive atmosphere are also reported [297,298].
AM technologies also have the potential to produce a variety of TMCs with hybrid reinforcements by the in-situ reactions.TMCs with TiB and TiC reinforcements can be fabricated using the additive B 4 C, wherein the in-situ reaction of 5Ti + B 4 C → 4TiB + TiC takes place during the AM process [299].According to Han et al [299], incorporating 1 wt% of B 4 C enhances the strength of the (TiB + TiC)/Ti composite fabricated via LPBF.However, the fracture strain of the composite is also significantly lowered.With an increased addition of B 4 C, the microstructure of the fabricated composites undergoes a transition from a lath-shaped structure to a cellular and dendritic structure [299].Therefore, the composites experience a strengthening effect from both dispersion strengthening and grain refinement strengthening.TMCs with TiN and Ti 5 Si 3 reinforcements can also be fabricated via the in-situ reaction of 9Ti + Si 3 N 4 → 4TiN + Ti 5 Si 3 .This method was employed by Gu et al [300,301] to fabricate (TiC + Ti 5 Si 3 )/Ti composites and the resulting composites demonstrated significant improvements in hardness and wear resistance compared to pure Ti.Similarly, (TiC + Ti 5 Si 3 )/Ti composite can be produced by the addition of SiC [215].However, it is important to consider various factors, including the reinforcement content, starting powder, and different principles of AM technologies, while design and production of TMCs bin the AM process.The mechanical properties of some AM-produced TMCs are summarized in table 6.
Indeed, some common laws/mechanisms can be summarized from the data in table 6.The increased reinforcement (or additive) content would increase the hardness and strength of the synthesized composites.Besides the dispersion strengthening, other strengthening mechanisms can also be introduced by adding reinforcements.For example, improving the metal matrix via the inclusion of TiB 2 into Ti may be ascribed to the increased nucleation rate of the matrix, which arises from the presence of reinforcements.As a result, additively manufactured MMCs often exhibit the Hall-Petch strengthening.This effect is due to the presence of reinforcements that may enhance the nucleation rate of the matrix.The Hall-Petch relationship could be used to estimate the Hall-Petch strengthening in MMCs [319][320][321][322]: where σ H is the yield strength of metal matrix with refined grains, σ o is the yield strength of a single grain for a metal matrix, k is the strengthening constant, and d is the average grain size of the metal matrix.The strengthening effect could also be written as: For instance, Jin et al [323] employed the LPBF technique to fabricate TiB/Ti composites with the TiB 2 content of 0-7.5 wt% based on the in-situ reaction.As depicted in figures 11(a)-(d), the average grain width of α ′ -Ti reduces from 1.03 µm to 0.36 µm as the TiB 2 content increases.Such refinement in grains is attributed to the increased TiB 2 content.The augmented TiB 2 content provides more nucleation sites, thus impeding the growth of α ′ -Ti grains in the LPBF process.Consequently, the hardness of TiB/Ti composites is enhanced from (258 ± 10.3) HV to (435 ± 14.7) HV (figure 11(e)).As multiple strengthening mechanisms operate concurrently, the correlation between the hardness of composites consisting of TiB/Ti and the average grain size typically follows the Hall-Petch relationship.However, it should be noted that the incorporation of reinforcements is observed to markedly elevate the viscosity of the melt, thereby impeding fluid flow and diminishing the rheological performance of the melt pools.As such, the pores between the particles may not be permeated and left in the produced parts.A tradeoff between grain refinement and increased viscosity necessitates further in-depth investigation.By comparing the outcomes in [302,303], it is found that the TiB/Ti composite produced by LPBF exhibits elevated strength levels, albeit with a concomitant reduction in strain, as compared to their counterparts manufactured by DED, despite having identical reinforcement content.The variance in mechanical properties can be ascribed to the principles underlying LPBF and DED.Figures 12(a) and (b) show that the fastcooling rate of materials during the LPBF process contributes to a markedly refined microstructure in the TiB/Ti composite.While DED also utilizes a laser as the heat source, the cooling rate of materials during the process is relatively lower than that in the LPFB process.Hence, the sizes of Ti and TiB grains in the DED-produced TiB/Ti composite are relatively larger than those in the LPBF-produced TiB/Ti composite (figures 12(c) and (d)).A similar scenario is also found in TiC/Ti6Al4V composites [311,312,316].Borisov et al [316] added 0.45 vol% TiC in the LPBF-produced TiC/Ti6Al4V composite.The composite exhibits a tensile strength of 1143 MPa.Such a result is comparable to the tensile strength of DED-produced TiC/Ti6Al4V composite with a TiC content of 10 vol% [312].Therefore, it can be concluded that even with the identical types and contents of reinforcements in the composites, AM technologies lead to discernible differences in their respective properties.Due to the different principles in various AM    technologies, the characteristics, such as output power, beam spot size, melt pool size, temperature gradient, and material cooling rate during the process, are significantly different.The characteristics of AM technologies are a prerequisite for producing high-performance MMCs and optimizing processing parameters.

Nickel matrix composites (NMCs)
Ni and Ni alloys are extensively used in turbines and petrochemical plants, making hard reinforcements with a nickel matrix highly promising for high-temperature structural materials [324,325].Many Ni alloys are successfully produced by AM methods, such as Inconel 625 [326], Inconel 718 [327], and Ni-based superalloys [328,329].Therefore, using AM methods to fabricate NMCs has aroused great interest among researchers.Both PBF and DED methods can fabricate NMCs [330,331].Various reinforcements, such as CNTs [330], TiC [331], WC [332], and SiC [329], have been used in the additively manufactured NMCs via external addition or in-situ synthesis.
Hong et al [331] fabricated TiC/Inconel 718 composites by LPBF using mixed powder and demonstrated that the Ni grains are refined by adding TiC.TiC particle and Ni matrix can form coherent interfacial layers with a width of 0.8-1.4µm.
Due to the refinement of Ni grains and stable reinforcementmatrix interfaces, the NMCs exhibit enhanced wear and tensile properties.However, if the input energy exceeds a certain threshold (160 kJ•m −1 ), the coarsening of dendrites in the NMCs could degrade their wear and tensile properties.Wang and Shi [333] conducted a similar work.They mixed 0.5 wt% TiC particles and Inconel 718 powder as feedstock and prepared Inconel 718 matrix composite by LPBF.Afterward, the prepared NMCs were solution-treated at 980 • C and 1100 • C and then aged at 620 • C and 720 • C. The results indicated that nano-particles are identified as effective reinforcements for the metal matrix, exhibiting enhanced strength in both as-built and heat-treated states.Additionally, the presence of nano-particles is found to be crucial in refining the microstructure of the Inconel 718 composite at temperatures below 980 • C.Under the 980 • C + aging condition, a maximum tensile strength of 1370 MPa is achieved, indicating a 16% improvement compared to the unreinforced TiC/Inconel 718.
Besides LPBF-produced NMCs, DED-fabricated NMCs have also been reported.Promakhov et al [334] prepared TiB 2 /Inconel 625 composites using mixed powders of NiTi-TB 2 and Inconel 625.The content of NiTi-TB 2 is in the range of 5 wt% to 100%.They proposed that DED is a promising technique for introducing less than 5 wt% NiTi-TB 2 in the Inconel 625 composite.When the content of reinforcements is over 70 wt%, the high internal stresses would cause the brittle fracturing of built samples.Hong et al [335] found that if the laser energy input is in the range of 80-120 kJ•m −1 in the preparation of Inconel 718/TiC, a coherent interfacial layer (identified as (Ti,M)C, M = Nb or Mo), with a thickness ranging from 0.8 µm to 1.4 µm, is established between TiC particles and the matrix.The formation of an interfacial layer enhances the hardness and wear performance of TiC/Inconel 718 composites.Gu et al [336] prepared TiC/Inconel 718 composites using different sizes of TiC particles, and they found that incorporating nano-TiC particles leads to the formation of refined columnar dendrites with well-developed secondary arms.Conversely, adding micro-TiC particles results in coarser and highly degenerated columnar dendrites, suppressing secondary dendrite growth.Therefore, the nano-TiC/Inconel 718 composite shows better hardness, tensile strength, and wear resistance without compromising ductility.Table 7 summaries the properties of some NMCs produced by AM technologies.

Iron matrix composites (IMCs)
IMCs exhibit high strength, stiffness, modulus, wear resistance, fatigue resistance, and corrosion resistance and, therefore, have been widely acknowledged in modern industries [345,346].IMCs produced by conventional technologies (such as infiltration casting [347] and powder metallurgy [348]) usually have poor wettability between the matrix and reinforcements, resulting in large residual stress, low density, poor strength, high cracking susceptibility, and reduced mechanical performance.Significant efforts have been made to use AM technologies to address these issues.
For example, Chen et al [349,350] have effectively produced WC-reinforced 1.276 7L tool steels containing different weight percentages of WC by LPBF using balling-mixed spherical powder.They observed that the martensite start temperature decreases with an increase in the weight percent of WC, which is attributed to the dissolution of W and C in the Fe matrix during the LPBF process.The in-situ reaction between WC and Fe is known to form (Fe,W) 6 C, significantly refining the resulting composite grains.Upon comparing the unreinforced sample, the composite reinforced by 2 wt% WC demonstrates a synergetic effect characterized by ultimate tensile strength of 1677 MPa, elongation of 8.5%, compressive strength of 3210 MPa, and fracture strain of 30.2%.Such an improvement in the mechanical characteristics can be ascribed to the combined influence of refinement strengthening, dispersion strengthening, and the effect of transformation-induced plasticity.
Riquelme et al [351] manufactured SiC/316L composites by DED with varying SiC contents up to 80 wt% and process parameters.The SiC reacts with molten 316L steel, producing Cr and Fe carbides, which results in the increased hardness of the composites.The highest hardness of 1085 HV is achieved for the 40 wt% SiC-addition composites.However, the increased weight percentage of SiC results in forming brittle and large-sized graphite precipitates, decreasing the hardness of the SiC/316L composites.Table 8 summarizes the mechanical properties of some AM-produced IMCs.
Moreover, Putra et al [358] synthesized a mixed Feakermanite printing ink with varying akermanite contents and fabricated porous Fe-akermanite composite scaffolds by the extrusion-based AM method.Despite the differences in shape between the iron powder and akermanite powder, a water-based binder composed of a 5 wt% aqueous solution of hydroxypropyl methylcellulose is used to mix the powder combination and produce the printing ink (figures 13(a)-(c)).The green parts that are produced are sintered to remove the printing ink.Small akermanite particles with irregular shapes are dispersed across the struts and adhered to the iron powder (figures 13(d)-(i)).Adding akermanite results in the improved adhesion of mouse embryonic osteoblastic precursor cells (MC3T3-E1) and high levels of cell proliferation.

Others
AM technologies can also produce other MMCs, copper matrix composites [359], and cobalt matrix composites [274][275][276].Many reports can also be found on Cu matrix composites.In recent decades, Cu matrix composites have garnered significant attention because of their good mechanical properties and high electrical/thermal conductivity, making them ideal for lightweight macroscopic conductors in electronics [360].AM technologies can also be used to produce Cu matrix composites.
In this regard, Constantin et al [361] have successfully developed diamond/Cu composites with low porosity by integrating recoating and remelting steps into the conventional LPBF.Recoating and remelting are two key strategies employed in the AM process to improve the quality of printed parts.Recoating involves passing a recoater over the previously printed layer to eliminate pores or voids.On the other hand, remelting involves melting the thin layer of powder deposited by the recoating process to fill any remaining pores and suppress printing defects.As depicted in figure 14(a), it is evident that the utilization of the recoating process results in increasingly pronounced porous surfaces with higher scan speeds and larger hatch distances.Cracks are found in the reduced scan speed and hatch distances due to critical thermal gradients being generated [362].A similar trend is observed using the remelting strategy (figure 14(b)).However, producing dense diamond/Cu composites requires a narrow processing window.Figures 14(c) and (d) show that the remelted Table 7. LPBF is laser powder bed fusion, and DED is direct energy deposition.The superscript T indicates the results of the tensile and compressive tests, respectively.samples have higher relative densities than those produced using other strategies.The printing, recoating, or remelting strategies for producing different sample surfaces are depicted in figures 14(e) and (f).However, it should be noted that the conventional printing and recoating strategies still result in rough surfaces of samples during the preparation, which may cause the laser beam to shift focus and affect its accuracy in melting the powder [363].Therefore, remelting is a good option for producing diamond/Cu composites.Metals, such as bronze, Al, and Co, could also be used as infiltrates in BJ printing [364][365][366].Molten metals can fill the porosity of the reinforcement preform with minimal shrinkage and less grain growth.Cui et al [364] used BJ printing to produce 316SS, 420SS, and WC parts, and melted bronze was used as infiltration to produce composites.The preform particles are evenly distributed and surrounded by a bronze matrix, as shown in figure 15.No interfacial layers are observed in the three composites.However, some pores are still found in the sharp edges of particles.The resulting composites, 316SS/bronze, 420SS/bronze, and WC/bronze, have porosities of (2.2 ± 0.3)%, (1.6 ± 1.4)%, and (1.7 ± 1.4)%, respectively.Mariani et al [367] produced WC-12Co composites using BJ printing and employed sintering/sintering plus HIP to densify the produced composites.The densities of the produced samples are 97.4% and 99.3% after sintering and sintering associated with HIP, respectively.Layeroriented pores are only found in the printed and sintered samples.Therefore, HIP is a useful method for reducing the porosity of BJ-fabricated composites.It is well known that the BJ-fabricated parts always have high porosities, regardless of composites or alloys [104,367].Therefore, post-treatments are often employed.Additionally, BJ-fabricated composite parts often have poor surface finishes and low accuracies, leading to a significantly small percentage of AM-produced composite parts being BJ-fabricated [104].

Mechanical properties of additively
The ultimate tensile strength and fracture strain of MMCs are summarized in figure 16.Conventional MMCs exhibit moderate strength and fracture strain.AM techniques offer merits in terms of fabrication speed, design flexibility, and material properties for alloys and MMCs [47,48,127].However, the distinctive processing procedure of AMproduced MMCs leads to different microstructures and mechanical performances compared to conventional MMCs.For instance, AM-produced TMCs show higher ultimate tensile stress than conventional TMCs due to the needle-like martensite α ′ produced in the microstructure during some AM processes (e.g.LPBF).The needle-like martensite α ′ does not contribute to the ductility of the Ti matrix, resulting in lower strain than conventional TMCs.The primary limitation of AM-produced TMCs is their ductility, which is sensitive to defects.AM has been shown to produce AMCs with high strength compared to their monolithic Al alloy counterparts, highlighting the potential of AM technologies to achieve exceptional mechanical properties for AMCs.AMCs also exhibit good ductility, making them suitable for various structural applications.The mechanical performances of AMproduced composites are comparable to those of conventional MMCs, enabling them to satisfy a wide range of mechanical specifications for advanced devices and applications.

Strengthening mechanisms
MMCs, whether produced through conventional methods or AM, typically exhibit higher strength than the corresponding metal matrix.This enhanced strength is a characteristic feature observed consistently in both conventional and advanced manufacturing techniques for MMCs.The high strength of MMCs could be ascribed to various strengthening mechanisms.In addition to the previously discussed Hall-Petch strengthening and dislocation strengthening, the enhancement of strength in MMCs produced via AM technologies is also attributed to the contribution of load transfer strengthening and Orowan strengthening mechanisms.Such strengthening mechanisms are introduced in detail as follows.

Load transfer strengthening
Load transfer strengthening is a crucial mechanism contributing to the higher strength of MMCs fabricated by AM  technologies and is well-described by the lag theory [368].The shear lag theory describes the ability of reinforcements in MMCs to transfer tensile stress from the matrix to reinforcements, employing effective shear stresses on the interfaces between the matrix and reinforcements.Load transfer strengthening is typically used in composites reinforced by fibers or reinforcements with non-uniform shapes [223].The load transfer strengthening could be described by the following expression: where σ cy is the yield strength of the composite, σ my is the yield strength of the matrix, V p and V m are the volume fractions of reinforcement and matrix, and R is the average aspect ratio of the fiber reinforcement.Hence, the strengthening effect (∆σ LT ) can also be expressed by:

Orowan strengthening
The Orowan mechanism represents significance in the reinforcement of materials when dislocations encounter obstacles in their glide planes, necessitating bypassing [369].The level of stress necessary for bypassing such obstacles indicates the Orowan strengthening.Such a strengthening mechanism is generally applied to the composites reinforced by particles, which is expressed by the following equation [369,370]: where ∆σ Or is the strength increment owing to the Orowan strengthening, M is the average orientation factor, r is the average particle diameter, ν is Poisson's ratio, λ is the average interparticle spacing, and other parameters are the same as mentioned above.

Sum of stress increment contribution
Substantially, the increased strength in MMCs is due to the synergistic effect of multiple strengthening factors.Sanaty-Zadeh [370] proposed that similar strengthening mechanisms can be combined using the root of the sum of the squares.Therefore, an equation suggested by Clyne as follows [370]: where σ y and σ y0 are the yield strength of the reinforced and unreinforced matrix, and the second term on the right is the The components are fabricated using a layer-by-layer method during the AM process, such as PBF, DED, and thermal spraying additive manufacturing, providing multiple thermal cycling of materials.
The influence of this enhancement becomes pronounced with the increase in temperature due to the growing disparity in expansion and contraction rates between matrix and reinforcements.This mechanism is less relevant at lower temperatures and can be disregarded at room temperature.
Load transfer strengthening The reinforcements transfer tensile stress from the matrix to reinforcements, employing effective shear stresses on their interfaces.
AM technologies promote the uniform distribution of reinforcements in the metal matrix, providing more matrix interface areas and reinforcements.
This mechanism is substantially reliant on the interface quality between the metal matrix and the reinforcements, with reinforcements generated through in-situ reactions demonstrating more stable interfaces than ex-situ counterparts.

Orowan strengthening
The magnitude of stress that is necessary for dislocations to bypass the reinforcements.
AM technologies promote the uniform distribution of reinforcements in a metal matrix, providing a greater dispersion-strengthening effect.
This mechanism assumes significance only when the size of the reinforcements is less than 1 µm total increment in the yield strength of MMCs.∆σ g , ∆σ d , ∆σ LT , and ∆σ Or are the increment contributions of refinement strengthening, dislocation strengthening, load transfer strengthening, and Orowan strengthening in the yield strength, respectively.The method above primarily applies to the reinforcements or the matrix grains at the micrometer scale, as it may not hold for nanocomposites.This phenomenon arises because the properties and performance of composites and alloys can significantly vary as they approach the nanoscale [371][372][373][374][375]. Table 9 summarizes the information on the strengthening mechanism in AM-produced MMCs.Generally, the contribution to the mechanical properties by AM technologies can be attributed to the promotion of uniform distribution of reinforcements, multiple thermal cycles of materials, and high cooling rate during the process.

New methods and new technologies for investigating AM-produced MMCs
The design and preparation of high-performance MMCs pose significant challenges.Each step demands scrutiny, from the meticulous selection of reinforcements and metal matrices to the thoughtful consideration of the AM process.Despite meticulous planning, failures in the fabrication of MMCs are not uncommon.The intricate interplay of various factors, such as material compatibility, processing parameters, and reinforcement distribution, requires a nuanced understanding to achieve the desired performance characteristics.The design of MMCs has conventionally been approached using a singlestep method, as depicted in figure 17(a).This linear method involves a series of sequential steps, each based on the results of the preceding step.However, because of its time-consuming and expensive nature, there has been a shift towards developing high-throughput methods to overcome the limitations of the single-step method.High-throughput methods facilitate the rapid optimization of the composition, process, and performance of MMCs and are often based on the material database.Compared to the single-step method, the highthroughput method offers several advantages, such as parallelization, as illustrated in figure 17(b) [376].
The design and preparation of MMCs are complex because of the many variables involved.The design of MMCs involves several variables, such as the type, content, size, and distribution of reinforcements, as well as the characteristics of the metal matrix and the features of the interface between the matrix and reinforcement.Although these variables offer flexibility in the design and preparation of MMCs, they also bring complexity to the process.High-throughput preparation methods are employed to overcome this complexity, which involves fabricating many samples covering all the variables needed.As reported, laser AM technologies (e.g.LPBF, laser cladding, direct laser deposition) are useful in producing alloys and composites using high-throughput methods [377].Shishkovsky et al [270] illustrated the use of laser AM in fabricating a graded layered TiB x /Ti-6Al-4V composite with different contents of TiB 2 addition.The particle content increases proportionally with the distance from the substrate, enabling the examination of microstructures and performances of MMCs with varying TiB 2 contents within a single composite.This approach eliminates the necessity for additional samples.Similarly, Kong et al [378] fabricated TiC/Inconel 718 composites using a high-throughput dual-feed laser metal deposition system, which allows for the fabrication of samples with different TiC contents (from 0 wt% to 10 wt%) on a single plate (figure 18(a)).TiC/composites with different TiC contents can be fabricated in a batch (figure 18(b)).Other investigations using similar techniques can be found in [379,380].AM technologies are suitable for preparing mass MMC samples with different variables, followed by high-throughput characterization technologies to obtain information on their compositions, microstructure, performances, and interfaces.Overall, the high-throughput methods give a powerful solution to designing and fabricating MMCs with tailored properties [381,382].
A profound comprehension of the processingmicrostructure-property relationship of AM-produced MMCs necessitates further research, technological advancements, and enhancements in materials and methods.The utilization of in-situ technologies, such as aberration-corrected TEM and synchrotron radiation x-ray microscopy, can offer valuable insights into defect generation, reinforcement distribution, and microstructural evolution [316][317][318].For instance, Liu et al [317] employed synchrotron radiation x-ray microscopy to investigate crack initiation and propagation in carbon fiber/ epoxy composites during loading, revealing nuanced nanoscale mechanisms that are not discernible through traditional methods.On the numerical simulation front, exploring the metallurgical behavior of multiphase materials in the molten pool during the AM process can provide insights into energy, momentum, and mass transmissions, leading to the development of corresponding numerical models [319].Additionally, implementing an AM monitoring system is essential for realtime data collection during the AM process.These challenges may be effectively addressed with cutting-edge technologies and numerical simulation techniques.

Intrinsic nature of MMCs coupled with AM technologies
The uniform distribution of reinforcements within MMCs is essential for realizing their intended performance.The strategic addition of the appropriate material to specific locations plays a vital role in manufacturing MMCs.Traditional manufacturing techniques, especially those involving liquidbased MMC production methods, encounter difficulties ensuring a completely homogeneous distribution of reinforcements within the matrix [383].In contrast, AM technologies possess the potential to offer a uniform and tailored distribution of reinforcements within the metal matrix [132,384].This advantage can be realized through proper feedstock preparation, as discussed in section 3.As outlined in table 2, mainstream preparation methods of composite feedstock, especially the pre-alloyed powder, can lead to a relatively homogeneous distribution of reinforcements (or additives) in the metal matrix.Despite the mechanical blending method exhibiting a relatively lower level of homogeneously distributed reinforcements, it remains superior to liquid-based techniques [245,383].Even in scenarios where the composite feedstock is not prepared before the AM process, the DED process enables reinforcements to be introduced into the melt pool of the metal matrix through separate nozzles, allowing for precise control over the amount of reinforcements [309,310].However, such merit is still not achieved in the PBF process using mechanically blended powder as feedstock.Therefore, despite the advantages that AM technologies offer in producing MMCs with a uniform distribution of reinforcements, it is acknowledged that pre-alloyed feedstock remains the optimal approach for ensuring uniformly distributed reinforcements within MMCs.However, it should be noted that the production of pre-alloyed feedstock is still associated with significant costs.There is an urgent need to develop low-cost pre-alloyed feedstock to address this challenge.
Although AM technologies provide many advantages for producing MMCs, they still have limitations.In contrast to conventional manufacturing techniques, AM technologies are associated with high costs, limited availability of feedstocks, constraints on the production of large-sized components, anisotropic mechanical properties, and challenges in achieving mass production.Acknowledging that these limitations are inherent to AM technologies, regardless of whether the output involves composites or alloys, is crucial.Consequently, the advancements in AM-produced MMCs are intricately linked to progress in AM technologies.For instance, the customization of MMC designs might require a specific feedstock that is not readily available in the market.
Furthermore, methods such as LPBF and EBPBF, known for their high resolution, come with elevated processing costs and extended processing times.Additionally, the size restrictions of processing chambers in these methods limit the dimensions of the produced parts.In the context of conventional parts, the primary challenges associated with AM technologies are centered on high costs and prolonged processing durations, thereby restricting the potential for mass production [385].Fortunately, when fabricating customized MMC components with complex geometries, such as bionic, topology-optimized, metamaterial, and superstructure designs, AM technologies can prove cost-effective [386][387][388].

Challenges in the AM processes
The potential for broad utilization and a bright future of SLM technology is contingent on producing parts without defects.The four AM processes mentioned in this article, PBF, DED, BJ, and spraying technologies, inevitably produce metallurgical defects in the MMC parts, e.g.balling, porosity, and cracking.Current research efforts have been rigorously directed toward developing solutions; however, the issues in question remain unresolved.The balling effect, frequently observed in PBF and DED processes and arising from disrupted melt tracks or molten metal splatter, detrimentally affects surface quality and hinders the homogeneous deposition of new powder layers [389,390].This phenomenon is partly influenced by the methods used for powder preparation, which determine the quality of the initial powder and are predominantly governed by wetting properties [20].The balling effect is exacerbated in AM-produced MMCs, as the reinforcements typically necessitate distinct processing parameters from those of the matrix metals, attributable to changes in physical properties.Hence, the selection of processing parameters should carefully consider the physical properties of the materials.
Porosity, a common feature and disadvantage in all AMproduced parts, originates from inadequate melting and entrapment of gases within the melt pools in PBF, DED, and spraying technologies and the stacking of particles in BJ and spraying technologies [90,122,135].The instabilities in melt pool fluid flow can result in inadequate melting and gas entrapment, which can be mitigated by optimizing the processing parameters of the AM processes [391].Pore formation, attributed to particle stacking, can be alleviated by employing optimally designed feedstock and subsequent post-treatment procedures [392,393].Additionally, the incorporation of reinforcements in AM-fabricated MMCs tends to intensify porosity due to an enhanced balling effect.Dadbakhsh and Hao [394] found that porosity becomes more acute in AlSi10Mg alloys with Fe 2 O 3 reinforcement, particularly at higher reinforcement volume percentages.Unfortunately, although many works report that nearly compact parts can be produced (porosity ⩽ 0.5%) [161,395,396], full compact parts are rarely reported.
Cracking in AM processes can be broadly classified into hot and cold types.Hot cracks, also called solidification cracks, emerge during the late stage of solidification as grains develop continuous skeletons.Research indicates that fine equiaxed semi-solid structures can mitigate the amount of entrapped liquid, thereby facilitating grain rotation and deformation compared to dendritic structures [397].This phenomenon absorbs strain within the semi-solid state, thereby averting the onset and propagation of cracks.Introducing reinforcements that act as nucleation sites promotes the transition from dendritic to equiaxed structures [261,398].The cold crack, attributed to residual stresses, is more significant in the AM processes [399].These stresses can cause cracks and separation between layers of material [226].In materials with multiple phases, internal strains develop from the differing coefficients of thermal expansion among the phases [400].One way to manage residual stress is by pre-heating the base substrate, which can reduce the temperature gradient.For MMCs made using AM, added reinforcements can increase the viscosity of the melt pools, affecting their flow.A greater energy density input is often required during manufacturing to distribute the particles evenly and prevent clumping.A high temperature is beneficial because it reduces the negative effects of viscosity on the spreading of the liquid material, which helps create a dense final product.However, introducing high-energy input and temperature gradients have been shown to induce cracks due to residual stress [401].This phenomenon creates a delicate balance between achieving fine grains and managing viscosity, which needs more in-depth research for better understanding and control.
As mentioned above, the defects in the MMC parts produced during the AM process would degrade their performance.Therefore, MMCs produced using the same AM process and with similar reinforcement content may demonstrate diverse properties, primarily due to variations in processing parameters [211,243,310,311].Hence, identifying optimized parameter sets to ensure the reproducibility of properties remains a significant concern in AM-produced MMCs.Many AM-produced MMCs involve in-situ reactions [36].In-situ reactions between additives and the metal matrix during the AM process may involve endothermic or exothermic reactions, requiring careful consideration during parameter optimization.
Additionally, specific AM processes, such as LPBF, involve rapid cooling, and the incorporation of ceramic reinforcements may render MMCs susceptible to cracking.Thoughtful consideration is necessary when selecting the types, morphologies, and quantities of reinforcements, considering potential in-situ reactions and the susceptibility of the metal matrix to cracking.Therefore, the processing parameters should be optimized to avoid stress concentration and improve the performance of the MMCs that are produced.As such, the optimization of parameter sets is particularly crucial for in-situ formed reinforcements, where the use of appropriate parameters during the AM process is essential.
Unfortunately, no existing database currently provides comprehensive information regarding processing parameters to guide the manufacturing of AM-produced MMCs.Future advances in the AM processes are still a challenge.

Future interests
In the future, two important research directions for the AMproduced MMCs should be significantly considered.One is the optimization of the cost of AM technologies.The other is the potential applications of AM-produced MMCs.The economic aspect of MMCs fabrication via AM often surpasses conventional methods.Present investigations into the costs associated with AM suggest that the technology is financially viable when producing limited quantities within centralized manufacturing settings [402].Nonetheless, as automation advances, the potential for cost-effectiveness in distributed production surges correspondingly.The costs of AM technology currently consist of material costs, machine costs, build envelope utilization, build time, energy consumption, and labor costs [402].For a rough estimation, material costs take about 30%-50%, depending on the type and quantity of materials used.The costs of machines and energy are about 10%-20% and 5%-15% [403,404].The combined influence of these factors makes cost reduction in AM technologies challenging and complex.Accordingly, while a thorough evaluation of cost reduction encompasses all factors throughout the AM process, selecting and optimizing materials may constitute a significant portion of the overall expenses [404].
Combining the advantages of AM technologies, such as design freedom, material savings, topologically optimized design, rapid prototyping, customized production, and reduced assembly [39], and the merits of MMCs, such as improving strength and rigidity, reducing component weight, improving wear resistance and corrosion resistance, improving thermal performance, and improving fatigue life [1,405], AMproduced MMCs have possessed considerable prospects for potential application in many industrial sectors.For example, AM-produced AMCs and TMCs can be used in the aerospace field, such as engine components, fuel nozzles, lightweight structural frames, space capsule interior structures, mars base building parts, and space suit accessories owing to their high specific strength and personalized customization [68,284].Because of their high strength, compatibility, and complex geometries, AM-produced TMCs can also be used in the biomedical field, such as custom orthodontics, orthopedic implants, and surgical tools [293].Due to their excellent hightemperature performance, corrosion resistance, and oxidation resistance, AM-produced NMCs are widely used in harsh working environments, such as structural materials for nuclear reactors, turbochargers and exhaust systems in vehicles, valves and pumps in chemical and petroleum industries [330,338].Therefore, the potential application of MMCs in AM is broadly acknowledged, particularly concerning advancements in aerospace and automotive engineering that necessitate enhanced performance and mass efficiency.Technological advances have made producing functionally graded materials a reality, further driving the optimization of material properties.Research is focused on enhancing the key physical attributes of AM-produced MMCs through process innovation and expanding their applications in personalized manufacturing and repair.Interdisciplinary collaboration is essential for the development in this field, and future research will focus on the demands of cost effectiveness and efficient resource utilization.

Conclusions
This article presents an overview of the latest advancements in AM of MMCs.These developments include AM technologies, feedstock, reinforcements, MMCs, and strengthening mechanisms.Due to their exceptional properties, MMCs have been used for numerous years, finding applications across diverse industrial sectors.However, the traditional techniques employed in manufacturing MMC components are cost-intensive and characterized by several processing steps.Consequently, AM techniques have emerged as a promising route for producing MMCs.Although multiple AM techniques have been developed in the last few decades, powder-based AM techniques (including PBF, DED, BJ, and spraying technologies) remain prevalent for the fabrication of MMCs.
This article discusses the significance of feedstocks in the performances of additively manufactured MMCs.The preparation of feedstocks is usually accomplished through in-situ and ex-situ methods.The former produces a composite powder with a homogeneous distribution of metal matrix and reinforcements, while the latter produces a mixture of metal matrix powder and additives.In any case, achieving a uniform distribution of reinforcements is crucial to producing highperformance MMCs.Reinforcements with high thermal stability, hardness, and compatibility with the metallic matrix should be selected for optimal results.This review highlights that oxides, carbides, and borides are commonly used as reinforcements in MMCs.
The AM technologies have successfully fabricated highperformance MMCs such as Ti, Al, Fe, and Ni-based MMCs.Compared to the pure metal matrix, the incorporation of reinforcements in MMCs refines their microstructures and prevents the production and propagation of cracks, thereby leading to improved mechanical properties.The fast-cooling rates during the AM process further refine the microstructures of MMCs.Therefore, the resulting AM-produced MMCs exhibit the advantages of both AM technologies and reinforcements.Nevertheless, much work must be done before the widespread commercial applications of AM-produced MMCs can be realized.Specifically, a comprehensive understanding of the processing-microstructure-properties of AMproduced MMCs is needed to guide their design, processing, and optimization.

Figure 2 .
Figure 2. Illustration of the generation of TiB particles via the in-situ reaction between Ti matrix and TiB 2 : (a) TiB 2 particles surrounding the large Ti powder in the powder mixture (b) in-situ formed TiB, semi-reacted TiB, and unreacted TiB 2 particles [143].[8 January 2017], reprinted by permission of the publisher (Taylor & Francis Ltd, www.tandfonline.com).

Figure 3 .
Figure 3. Phase constituent, relative densities and properties of Al matrix composites with varied sizes of SiC particles produced by laser powder bed fusion: (a) XRD patterns, (b) relative densities, (c) hardness, and (d) wear rates.Reprinted from [213], Copyright © 2015 Elsevier B.V. All rights reserved.

Figure 4 .
Figure 4. Microstructural features of (ZrC/TiC)/Al composites produced using laser powder bed fusion: (a) and (b) are SEM images, (c), (d), and (e) are EDS mapping of Ti, Zr, Al in (b); (f), (g), (i), and (j) are TEM images of TiC and ZrC particles in the Al matrix; the HAADF images of interfaces as marked by rectangular are indicated using the insets in (g) and (j); (h) and (k) are EDS results of line scan in (g) and (j) along with the red arrows.Reprinted from [217], © 2022 The Authors.Published by Elsevier B.V.

Figure 5 .
Figure 5. Microstructure of laser powder bed fusion produced Al-Cu-Ag-Mg-Ti-TiB 2 sample: (a) and (b) inverse pole figure maps of Al and TiB 2 grains; (c) size distributions of Al and TiB 2 grains; (d)-(f) scanning transmission electron microscopy (STEM) in high-angle annular dark field (HAADF) and energy disperse spectroscopy results of Al-Cu-Ag-Mg-Ti-TiB 2 sample showing the distributions of Al 2 Cu and TiB 2 grains; (g)-(n) STEM in HAADF images of the interfaces of Al 2 Cu and TiB 2 with the Al matrix; the insets in (i) and (m) are fast Fourier transforms from Al, Al 2 Cu, and TiB 2 grains.Reproduced from [220].CC BY 4.0.

Figure 6 .
Figure 6.TEM images of microstructures of a TiN/AlSi10Mg composite produced by laser powder bed fusion: (a) bright-field images exhibiting the dislocation tangles in Al grains, (b) interactions of dislocations and TiN particles, (c) HRTEM image of the TiN/Al matrix interface, (d) HRTEM image of Mg2Si/Al matrix interface, (e) and (f) are HRTEM images of Mg2Si, Si, and Al phases with dislocations.Reprinted from [222], © 2020 Elsevier B.V. All rights reserved.
produced AlSi10Mg and TiB 2 /AlSi10Mg composite parts via cold spray AM and subjected them to post-friction stir processing.They found that eutectic Si particles and unevenly distributed TiB 2 clusters are distributed along the boundaries of Al grains (figures 8(a) and (b)).Fine Si nanoparticles are precipitated from the supersaturated Al matrix in a dispersed manner.Following post-friction stir processing, the Si particles are dispersed along Al grain boundaries and inside Al grains (figures 8(c) and (d)).Throughout the FSP, the material within the stir zone is subjected to intense deformation, generating a substantial density of dislocations.As the deformation persists, the energy accumulated within these dislocations propels the dynamic recovery and recrystallization.Combined with the dispersed Si particles, the grains in the matrix are refined.The TiB 2 clusters are fragmented in the post-treated parts, effectively improving ultimate tensile strength with values of 295 MPa for AlSi10Mg alloy and 365 MPa for TiB 2 /AlSi10Mg composite (figure 8(e)).The

Figure 7 .
Figure 7. Schematic illustration of sliding wear of (a) Al coating and (b) carbon nanotube reinforced Al composite coating.Reprinted from [227], © 2020 Elsevier B.V. All rights reserved.

Figure 8 .
Figure 8. Scanning electron microscopy images exhibiting the cross-sectional morphologies of the cold spraying and cold spraying + friction stir processing samples: (a) and (c) AlSi10Mg; (b) and (d) TiB 2 /AlSi10Mg composite; (e) tensile hardening and (f) strain hardening of the composites.CS denotes cold spraying, and FSP is friction stir processing.Reprinted from [229], © 2020 Elsevier Ltd.All rights reserved.

Figure 9 .
Figure 9. Interface between TiC and matrix in TiC/Ti composite produced by direct energy deposition: (a) image of the interface obtained by high-resolution transmission electron microscopy; (b) corresponding selected area electron diffraction of the interface.Reprinted from [286], © 2020 Elsevier B.V. All rights reserved.
is laser powder bed fusion, and DED is direct energy deposition.The superscripts T and C indicate the results of the tensile and compressive tests, respectively.

Figure 12 .
Figure 12.Comparison of microstructures of TiB 2 /Ti composites produced by laser powder bed fusion and direct energy deposition: (a) and (b) laser powder bed fusion produced TiB 2 /Ti composites.Reprinted from [302], Copyright © 2014 Acta Materialia Inc. Published by Elsevier Ltd.All rights reserved.(c) and (d) Direct energy deposition produced TiB 2 /Ti composites.Reprinted from [303], © 2017 Elsevier B.V. All rights reserved.
is laser powder bed fusion, and DED is direct energy deposition.The superscripts T and C indicate the results of the tensile and compressive tests, respectively.

Figure 13 .
Figure 13.Morphologies of the powder for the production of Fe-akermanite composite scaffolds by extrusion-based additive manufacturing method: (a) Fe and (b) akermanite; (c) the produced Fe-akermanite composite scaffolds have a 0 • and 90 • pattern; (d) the Fe-akermanite composite scaffolds; (e)-(h) the strut of Fe-akermanite composite scaffolds with different akermanite contents at high magnification; and (i) energy disperse spectroscopy mapping of Fe-akermanite struts.Reproduced from [358].CC BY 4.0.

Figure 16 .
Figure 16.Ultimate tensile strength and fraction strain of various composites produced by different methods.MMCs refer to metal matrix composites, TMCs indicate Ti matrix composites, AMCs mean Al matrix composites, and AMed is abbreviated for 'additively manufactured'.

Figure 17 .
Figure 17.Comparison of the processes in (a) the single-step method and (b) the high-throughput method.

Figure 18 .
Figure 18.The high-throughput dual-feed laser metal deposition system for fabricating Inconel 718 alloy with different TiC weight percentages: (a) schematic illustration of the system and (b) the fabricated samples.Reprinted from [378], © 2019 Elsevier B.V. All rights reserved.

Table 1 .
Different powder-based additive manufacturing technologies for fabricating metal matrix composites, the feedstock used, the processing mechanisms, advantages, and disadvantages.
Spraying technologyThe powder is sprayed at a high velocity; the powder is deformed and melted depending on the heat resource usedHigh production rate; it can be conducted at room temperature and in an atmospheric environment

Table 2 .
Various feedstocks used for producing metal matrix composites, methods, advantages, and disadvantages.

Table 3 .
Characteristics of widely used reinforcements or additives.

Table 4 .
Characteristics of frequently employed metal matrix materials.

Table 5 .
Mechanical properties of additively manufactured aluminum matrix composites.

Table 5 .
LPBF is laser powder bed fusion, DED is direct energy deposition, and CS is cold spraying.HT denotes heat treatment; the superscripts T and C indicate the results of the tensile and compressive tests, respectively.

Table 6 .
Mechanical properties of additively manufactured titanium matrix composites.

Table 8 .
Mechanical properties of additively manufactured nickel matrix composites.

Table 9 .
Summarized information on strengthening mechanisms in additive manufacturing produced metal matrix composites.