北京化工大学：《材料导论》课程阅读材料（金属材料）Fabrication_of_metal_matrix_composites_by_metal_injection_molding_review.pdf

journal of materials processing technology 200 (2008) 12–24 journal homepage: www.elsevier.com/locate/jmatprotec Review Fabrication of metal matrix composites by metal injection molding—A review Hezhou Ye a,∗, Xing Yang Liub, Hanping Hong a a Faculty of Engineering, The University of Western Ontario, London N6A 5B9, Canada b Advanced Materials Design, Industrial Materials Institute, National Research Council of Canada, London N6G 4X8, Canada article info Article history: Received 7 August 2007 Received in revised form 27 September 2007 Accepted 24 October 2007 Keywords: Powder injection molding (PIM) Metal injection molding (MIM) Metal matrix composite (MMC) abstract Metal injection molding (MIM) is a near net-shape manufacturing technology that is capable of mass production of complex parts cost-effectively. The unique features of the process make it an attractive route for the fabrication of metal matrix composite materials. In this paper, the status of the research and development in fabricating metal matrix composites by MIM is reviewed, with a major focus on material systems, fabrication methods, resulting material properties and microstructures. Also, limitations and needs of the technique in composite fabrication are presented in the literature. The full potential of MIM process for fabricating metal matrix composites is yet to be explored. Crown Copyright © 2007 Published by Elsevier B.V. All rights reserved. Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2. Metal matrix composites (MMCs) by MIM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.1. Refractory metal based MMCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2. Titanium based MMCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3. Intermetallics based MMCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.4. Steel based MMCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.5. Bimetal structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 3. MMC by microMIM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 4. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 ∗ Corresponding author. E-mail address: hye5@uwo.ca (H. Ye). 0924-0136/$ – see front matter. Crown Copyright © 2007 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2007.10.066

journal of materials processing technology 200 (2008) 12–24 13 1. Introduction Powder injection molding (PIM) is a near net-shape manufacturing technology that combines the shaping efficiency of plastic injection molding with the capability of powder metallurgy for processing metal and ceramic powders (German, 1990). The PIM process normally involves four steps: mixing, injection molding, debinding and sintering, as illustrated in Fig. 1. The evolution of the PIM technology has resulted in many variations, reflecting different combinations of powders, binders, molding techniques, debinding routes, and sintering practices. Metal injection molding, commonly known by its acronym MIM, is by far the most widely used PIM process. The process offers many unique advantages for the mass production of small and complex parts. First, precise and reproducible shapes can be achieved at a notably reduced cost as compared to conventional manufacturing processes. Part quantities varying from 5000 per year (e.g., specialty firearm sights) to over 100 million per year (e.g., cell phone vibrator weights) can be economically produced by MIM (Johnson and German, 2003, 2005). Second, complex shapes can be fabricated relatively easily with MIM, whereas conventional powder metallurgy methods are limited to the fabrication of parts with simple geometries. Third, the use of very fine powders in the feedstock promotes densification during sintering and hence high-performance parts can be produced with material systems that are difficult to sinter using conventional processes. The attractive features of the MIM process can be applied advantageously to the fabrication of metal matrix composite (MMC) or ceramic matrix composite (CMC) parts. Although many MMC or CMC offer unique properties that cannot be normally achieved by monolithic materials, their commercial use are often restricted by the high cost of materials and manufacturing. By applying the MIM process, the cost for commercial use of composite materials can be significantly reduced. In recent years, comprehensive work has been conducted to explore the potential of MIM for the fabrication of metaland ceramic-based composites and components. Companies have even been established to discover the commercial capacity of the MIM technology for the fabrication of composites (Decker, 1989; H.C. Starck Inc., 2003). The most widely studied composites by PIM are metal-based, including stainless steels, refractory metals, intermetallic compounds, and titanium alloys. Although theoretically the reinforcements in a composite can take either continuous (typically long fibres) or discontinuous (particles and short fibres/whiskers) form, the flow and mold filling requirements of the PIM process determine that it is most applicative for processing the feedstock containing particles or short fibres. Consequently, it is not surprising that all of the metallic composites fabricated by MIM to date are strengthened by discontinuous reinforcements. This paper provides a review of the research activities related to composite fabrication through the MIM route. For completeness, bimetal structures and surface-engineered components manufactured by MIM are also included in the metal matrix composites. Furthermore, fabrication of microcomponents by mircometal injection molding will be covered as well in the context even though it is still green at fundamental investigation stage. 2. Metal matrix composites (MMCs) by MIM Based on the matrix materials, the composites fabricated by MIM can be divided into refractory metal based MMC, titanium based, intermetallic based and steel based. Bimetal structures are also discussed in this section. The MIM route has enabled the fabrication of MMCs containing ingredient materials that are not compatible in molten state and difficult to fabricate by conventional routes (Diehl and Detlev, 1990). 2.1. Refractory metal based MMCs Tungsten and molybdenum are two refractory metals that have attracted great interest for high-temperature applicaFig. 1 – Schematic diagram of powder injection molding

14 journal of materials processing technology 200 (2008) 12–24 tions in such industries as consumer electronics, aerospace, telecommunications, medical and defence, due to their excellent heat resistance. The high-melting temperatures of the metals (3422 ◦C for tungsten and over 2623 ◦C for molybdenum), however, make it extremely difficult to process the materials by melting and casting. Conventionally, tungsten and molybdenum-based alloys and composites are fabricated by powder metallurgical method, followed by minor machining or grinding, if necessary. The components manufactured by the conventional powder metallurgy processes are normally limited to simple geometries. The shaping advantage of MIM makes it meritable for producing complex-shaped parts from tungsten or molybdenum-based materials. Therefore, MIM has been widely explored as an alternative manufacturing route for fabricating refractory metal-based materials (Bruhn and Terselius, 1999; Wang et al., 2001; Li et al., 2003; Tarata and Ghita, 2002; German, 1999; Bose, 2003; Yang et al., 2000; Sung et al., 1999). Currently, the fabrication of such refractory components represents one of the most profitable applications of PM technology (German, 1999). Due to the poor sinterability of the refractory system, ultrafine powders, particle size less than 10m or even at submicron scale, are commonly used in the fabrication of refractory parts by MIM in order to achieve high sintering density. A homogenous mixture of the powders should be attained prior to further processes. Mechanical alloying and the reduction of oxidized powders have been employed to get uniform mixtures of the starting materials (Moon et al., 1998a). The binder is a critical element in determining the part quality and the productivity of MIM process. New binder systems have been developed, some of which are listed in Table 1. As shown, wax-based binders are extensively used to formulate the MIM feedstock because waxes have low-melting temperature, good wetting, short molecular chain length, low viscosity and small volume change during thermal process. After injection molding, the green parts are subjected to debinding process, solvently or thermally, or both. In most cases, the debound refractory composite parts are sintered at over 1200 ◦C in pure hydrogen atmosphere. Final density, mechanical, thermal and electrical properties are the key parameters for the product (Moon et al., 1997). An important refractory metal-based composite is the W–Cu system, which has superior thermal management properties and high-microwave absorption capacity. It is predominantly used for heavy-duty electrical contacts and arcing resistant electrodes. W–Cu parts are usually fabricated by infiltrating Cu into W powder compacts or liquid-phase sintering of W/Cu powder mixtures. However, conventional techniques are unsuitable for the mass production of small intricate parts. MIM has been practised as an apposite method for mass production of small and delicately shaped W–Cu parts. In Yoo’s invention (Yoo et al., 1999), MIM is employed to fabricate a tungsten skeleton structure having the porosity in the range of 15–40% after sintering. Copper infiltration is then carried out by putting a copper plate beneath the tungsten skeleton structure and heating in hydrogen atmosphere. Compared to conventional P/M methods, the method can save energy and sidestep a sudden shrinkage during sintering. Furthermore, Yoo adopted the method for fabricating a hermetically sealed W–Cu package container for a microwave device and achieved a strip wire connection having the thermal expansion coefficient similar to that of GaAs without any extra machining process (Yoo et al., 2000). Knuwer (Knuwer et al., 2000) confirmed that using MIM followed by liquid-phase sintering was a promising and economical method for the production of housings for integrated HF-circuits. He obtained satisfactory accuracy as well as the mechanical and physical properties. Besides using ultrafine powders, adding a small amount of a third metal such as Co, Ni or Fe into the system was evidenced to improve the sinterability (Yang and German, 1993a). However, the introduction of the additive element compromises the thermal management properties of the composites, because purity and homogeneity control in theW–Cu system is critical for obtaining high thermal conductivity (Yang and German, 1994b; Yang and German, 1993b). The use of ultrafine powders, in particular nanosized W/Cu powders, provides an effective solution to the sinterability problem of the W–Cu system. W–Cu composites with nearly full density have been achieved (Moon et al., 1994, 1997, 1998a; Kim et al., 1992, 1999, 2000, 2006; Ryu et al., 1998; Yang and German, 1994a, 1997). For example, a W-20%Cu composite made of nanosized particles could be densified to 95% of theoretical density by sintering at 1150 ◦C for 2 h, and the W particles distributed homogenously in Cu (Moon et al., 1998a). To achieve uniform mixing of tungsten and copper powders and to increase the solid loading of nanosized powders, Kim and colleagues (Kim et al., 1999, 2000, 2006; Ryu et al., 1998) adopted low-energy ball milling to modify the shape of commercial Table 1 – Binder systems used in the feedstock of refractory composites for MIM Reference Composition Molding temperature Solid loading (vol.%) Debinding method Powder Moon et al. (1994, 1998a) 45% PW, 15% Bee wax, 30% PE, 10% SA 120–130 NA Thermal W–Cu Kim et al. (2000, 2006) Ryu et al. (1998) Kim et al. (1999) Yang and German (1997) 35% PP, 60% PW, 5% SA NA 52 Solvent + thermal W–Cu Yang and German (1994a) 40% PP, 55% PW, 5% SA NA 54 Solvent + thermal W–Cu Fan et al. (2005) PS, PP, EVA, PW, oil, SA, DOP 140–160 51 Solvent + thermal W–Ni–Fe Suri et al. (2003) PW, PP 170 60 NA 97W–2Ni–1Fe Fan et al. (2002) 74% PW, 20% EVA, 5% HDPE, 1% SA 135 57 NA 90W–7Ni–3Fe Huang et al. (2003) Li et al. (1998) 50% Polystyren, 30% PP, 20% VO 185 47 Solvent + thermal 97W–2Ni–1Fe

JOURNAL OF MATERIALS PROCESSING TECHNOLOGY 200(2008)12-24 fine powders.In the course of the low-energy ball milling sinterability is also an inherent problem.While at sintering the tungsten powder exhibited no change in either size temperature, there is essentially no solubility of Cu in M to a three MIM ny1-1.s7 y of M molte of the p owder shape and sulted in impro he fe ing ar s inves : Further high wders and prevent potential contaminations enc e densification.Kirk (Kirk et al 1992ds red tha tere inotherm ing processes. ant all min time to2h Undoubtedly.most of this impurity was ced from the stainless steel vial used for millin diffusivity.Using ultrafne powders seems to be more beneh alloyed W-Cu powder at 600C for 1h in a hydrogen atm thermaldiffusivity of the composite is to amplify thehigh con ents in the vity FCC pha by increasing the v Ryu et al.1998.2002:Moon et al 199b:Lee et al,2000).This ct Is to the rem 2.2 Titanium based MMCs osed Titanium alloys are the most metallic biomate to take advar ntage of different processes,such asg nding and rials for me dical implant appli ions due to their goo com posite powde for MIN used in aerospace applicat ions be ause o ca wide ngth and al.2002:Hu pimanyfocusedonappicationsinthebiomedicalfeld ue to the mbination of high der Thian and his colle agues are the leading rese the materials.The for MI posites for medical implant applications by MIM (Thia fabrication is9w- i-3Fe(Fan et al,2001ab, 021 005;Sur et a 2001.2002ab.c.2003).HAis calcium pr phate com H)2 le W Ni miinand mixed wit tive mat courages bone ingrowthsw dsto 1 natu 2001. the superic degree of mixing of th wders bioactitvity th ent n is of HA (Ha ani et al 1994 erties and negligible distortion were produced by solid- tat make a homoge neous feedstock T6A4 and HA powder I orde hol (PVA as the 03 The PVA y (Fan et al.,2005)propo d a mather natic model of adjustin rem d by hea atine.and the ush ight deviation of cted parts of additives like Mo.Ta a cture HA.A commercial binder syste PAN-250S anical propert syste y MI the experimer set appropriate ding an ndin Bose et al 19.1990).Their results show that vem al solid load quantity of Re ha solid solutio ngthenin ing of the composite powder in the fe fo mol the cations debound thermally and the heating rate and ga ar to the w-Cu composite Mo-C com mportant in the pro ha ely match to semiconductor mater The superiority of a higher rate at the later stage with a high gas flow rate demonstrated as an P

journal of materials processing technology 200 (2008) 12–24 15 fine powders. In the course of the low-energy ball milling, the tungsten powder exhibited no change in either size or shape, while the ductile Cu powder was deformed to a threedimensional equiaxial shape with reduced size. The change of the powder shape and size resulted in improved solid loading and better sinterability. Kim’s investigation indicated the low-energy ball milling could circumvent the agglomerate of nanopowders and prevent potential contaminations encountered in other milling processes. However, iron contamination was detected in the products even when the authors restricted ball milling time to 2 h. Undoubtedly, most of this impurity was introduced from the stainless steel vial used for milling and it could be minimized by using vials made from cemented carbides. Besides mixing and milling, reducing mechanically alloyed W–Cu powder at 600 ◦C for 1 h in a hydrogen atmosphere can also lead to improvements in the green density and shape stability of the powder compact (Kim et al., 2000; Ryu et al., 1998, 2002; Moon et al., 1998b; Lee et al., 2000). This effect is mainly attributed to the removal of internal impurities and oxides. In both Meinhardt’s and Lee’s inventions (Minhardt et al., 2002, 2005; Lee et al., 2005a), they proposed to take advantage of different processes, such as grinding and reducing oxides in hydrogen, mixing powders with a protecting liquid in an inert atmosphere, to make super fine and pure composite powders for MIM. Fabrication of W–Ni–Fe composites by MIM has been widely investigated as well (Fan et al., 2001a,b, 2002, 2004, 2005; Suri et al., 2003; Huang et al., 2003; Li et al., 1998; Lee et al., 2005b; Qu et al., 2000, 2001) due to the unique combination of high density, high strength, good ductility and corrosion resistance of the materials. The primarily studied W–Ni–Fe system for MIM fabrication is 90W–7Ni–3Fe (Fan et al., 2001a,b, 2002, 2005; Suri et al., 2003; Huang et al., 2003; Lee et al., 2005b; Qu et al., 2000, 2001). Nanoscale W–Ni–Fe composite powders are commonly fabricated by mechanical milling and mixed with a multicomponent binder to form homogeneous feedstock (Suri et al., 2003; Huang et al., 2003; Lee et al., 2005b; Fan et al., 2001a,b). Experimental results demonstrated that milling escalated the degree of mixing of the elemental powders, the maximum powder loading, homogeneity of the feedstock, and the final sintering. Fully densified parts with high-mechanical properties and negligible distortion were produced by solid-state sintering in the temperatures range of 1350–1450 ◦C. In order to control the dimensional deviation of the final products, Fan (Fan et al., 2005) proposed a mathematic model of adjusting the weight deviation of the injected green parts, which was verified by experimental investigation. Furthermore, the role of additives like Mo, Ta and Re on the microstructure and mechanical properties of the W–Ni–Fe system by MIM has been scrutinized by German and Bose (German et al., 1989; Bose et al., 1989, 1990). Their results show that even a small quantity of Re has remarkable solid solution strengthening and the grain refining capacity. However, the high cost of Re may restrict the addition of the additive to only special applications. Similar to the W–Cu composite, the Mo–Cu composite has high thermal conductivity and the thermal expansion closely match to semiconductor materials. The superiority of Mo–Cu to Wu–Cu is the lower density at the same weight fraction level. During the fabrication of the composite, poor sinterability is also an inherent problem. While at sintering temperature, there is essentially no solubility of Cu in Mo and only 1–1.5% solubility of Mo in molten Cu. Utilizing the MIM process to fabricate Mo–Cu composite improves sinterability due to the use of ultra fine particles in the feedstock. Further more, additive elements, having a higher solubility for Mo and forming a continuous solid solution with Cu, will promote densification. Kirk (Kirk et al., 1992) discovered that the addition of Ni to the systems could significantly enhance densification during sintering. However, this was achieved at the cost of coarsening Mo grain size, and abating thermal diffusivity. Using ultrafine powders seems to be more benefi- cial because it raises the thermal diffusivity of the composite together with densification. An alternative route to improve thermal diffusivity of the composite is to amplify the high conductivity FCC phase by increasing the volume of Cu (Nan et al., 2004). 2.2. Titanium based MMCs Titanium alloys are the most promising metallic biomaterials for medical implant applications due to their good biocompatibility, high-chemical stability in the physiological environment, and excellent mechanical properties. They are also extensively used in aerospace applications because of their high specific strength and corrosion resistance. The current research on titanium matrix composites through MIM is primarily focused on applications in the biomedical field. Thian and his colleagues are the leading researchers in exploring the fabrication of titanium–hydroxyapatite (HA) composites for medical implant applications by MIM (Thian et al., 2001, 2002a,b,c, 2003). HA is a calcium phosphate compound [Ca10(PO4)6(OH)2] with a composition similar to the mineral phase in natural bones. It is established as a bioactive material that encourages bone ingrowths when used in implants. HA itself, however, is brittle in nature, and cannot be used in load-bearing applications. The purpose of developing the HA/Ti alloy composite is to combine the superior bioactivity of HA and the excellent mechanical properties of titanium alloys, so as to avoid the brittleness and low fracture toughness problems of HA (Halouani et al., 1994). To make a homogeneous feedstock, Ti6Al4V and HA powders were first mixed by a slurry approach, using polyvinyl alcohol (PVA) as the carrier (Thian et al., 2003). The PVA was lately removed by heating, and the powder mixture was crushed into small particles to obtain Ti6Al4V/HA composite powders with an inner core of Ti6Al4V wrapped with an outer layer of HA. A commercial binder system, PAN-250S, was used in the experiment. To set appropriate molding and debinding temperatures, the thermal properties of the multi-component binder was investigated by TGA/DSC. The optimal solid loading of the composite powder in the feedstock for MIM (Thian et al., 2002a) was around 60 vol.%, essentially the same as most other kinds of feedstock for MIM. The molded parts were debound thermally and the heating rate and gas flow rate played an important role in the process (Thian et al., 2001). A slow heating rate at the beginning stage of debinding and a higher rate at the later stage, with a high gas flow rate during the whole process, were demonstrated as an effective approach to remove the binder

16 JOURNAL OF MATERIALS PROCESSING TECHNOLOGY 200 (2008)12-24 tering.a key challe nd noor fabricability of th in fabricating/HA composite is to pr nt HA fron ounds limit their applications.Composite strengthening decomposing.The degree of HA decomposition increases wit On t other hand,to achieve a high echniques have been veloped to produce the Ni,Al-base of the g ten off and A 90 wnat mak omposite with 50wt%of Ti and 50wt%of HA,a sinte Depending on composition,the strength of NigAl increase te hea nin and a c withtemperature up to microhardn ess.while temperature resulted ty of fabricating NiAl-based composite by MIM(Bose and n simulated。 ogical conditions revealed the complete sed in the experim ent.The preferential align ment of th lution of the secondary phase bers in the matrix was attained by a modified form of MIM hate (ICP).tet caophophae tals pre also foun interingofele nental powders under exter chan ted dur identical pre use of the p cipitation of ion thermal rocedure,and pressu as evidenced b MMCs by MIM Many issues.such as the nde ntenng mecha sms ot the stem,the mech Alman and toloff furthe expl the adaptability o al pro d to b Given sintered in controle man 1991;Alman et a han d an r im ents in the mechanical 1p03 an expecte HA-based com ynthesis foll d by hot ng.Results cate 2008 2002 testsre and wang. chanek et al,1997;Hoepfner and Case. tthe Al fbers would trengthen the ostrate,espe Oktar Weng et al.,1 a1g即ey al properties The rep rted/HA composite produc materaduetotshigh-meltnont()x ellent ox 50w de sintering.HA-HA and HA-TisAl4V would be bound togethe ever,hamper the manufacture and application of this reas the Ti6Al4V powde actually restricted der atio Adding vo of chopped for the high porosity of the composite 50%)in the ved fracture sistance and hardness (Stoloff ane tud Therefore the mechanical properti f the Alman,1991:Alm an et al..1992) te are d by the HA POW s an adv ting sh omposition of HA at lower temperature (Wang and Chal orientation of the hbre Research results show that the ng et al..1994) y n the p xpa e the will be able to advance the properties of the MIMed composite oduce a fibre alienment parallel to the flow direction thein Fig.2 9b).B molding,the mechanical pro rties along a c Ithasbeen recognized thatintermetallic compounds basedo an thus be tailored according to performance requirement w ae sity,hig which are hiehl y desirable for hig ittle effect on the degree of alignment (Alma temperature structural applications as rep owever,this met uffers from sev

16 journal of materials processing technology 200 (2008) 12–24 During the high temperature sintering, a key challenge in fabricating Ti6Al4V/HA composite is to prevent HA from decomposing. The degree of HA decomposition increases with sintering temperature, and it becomes significant when the sintering temperature reaches 1100 ◦C (Thian et al., 2002b). On the other hand, to achieve a high degree of densification of the composites requires a relatively high sintering temperature. Thian et al.’s results showed that for a Ti6Al4V/HA composite with 50 wt.% of Ti and 50 wt.% of HA, a sintering temperature of 1100 ◦C, a heating rate of 7.5 ◦C/min and a cooling rate of 5 ◦C/min produced the highest relative density and microhardness, while higher sintering temperature resulted in higher flexural strength and modulus. Furthermore, an in vitro study of the MIMed Ti6Al4V/HA tensile bars performed in simulated physiological conditions revealed the complete dissolution of the secondary phases such as tricalcium phosphate (TCP), tetracalcium phosphate (TTCP), and CaO after 2-week immersion. Following that, calcium phosphate crystals precipitated after 4 weeks of immersion. It was also found that the mechanical properties deteriorated during the initial immersion period and then gradually recovered to almost identical pre-immersion values because of the precipitation of an apatite layer (Thian et al., 2002c). It should be pointed out that the research on Ti/HA-based MMCs by MIM is relatively new. Many issues, such as the sintering mechanisms of the Ti/HA system, the mechanical properties of the composites, and the prevention of the decomposition of HA at high temperatures, need to be elucidated. Given that pure HA could be sintered in controlled conditions to achieve better mechanical properties than that reported for the Ti/HA composite (Halouani et al., 1994), further improvements in the mechanical properties of Ti/HA composites can be reasonably expected. Also, HA-based composites with such reinforcements as silver, silica, titania (Nair et al., 2008; Wang and Chaki, 1993; Gu et al., 2002; Chaki and Wang, 1994; Suchanek et al., 1997; Hoepfner and Case, 2003; Oktar, 2006; Chu et al., 2004; Weng et al., 1994) have also been sintered to get attractive mechanical and biological properties. The reported Ti6Al4V/HA composite produced by MIM consisted of 50 wt.% Ti6Al4V and 50 wt.% HA, with HA coated onto the Ti6Al4V powder in the feedstock. During sintering, HA–HA and HA–Ti6Al4V would be bound together, whereas the Ti6Al4V powder actually restricted densification of the composite (Thian et al., 2002c). That is one of the reasons for the high porosity of the composite (over 50%) in the study. Therefore, the mechanical properties of the composite are mainly contributed by the sintered HA powders and Ti6Al4V/HA interfaces. Also, sintering in vacuum would cause decomposition of HA at lower temperature (Wang and Chaki, 1993; Weng et al., 1994), especially with the presence of Ti. Altering the composition and revising the sintering conditions will be able to advance the properties of the MIMed composite. 2.3. Intermetallics based MMCs It has been recognized that intermetallic compounds based on aluminium have appealing characteristics of low density, high strength at elevated temperatures, along with good corrosion and oxidation resistance, which are highly desirable for high temperature structural applications as replacement for superalloys. However, the brittleness and poor fabricability of the compounds limit their applications. Composite strengthening offers a way to improve the mechanical properties. Among the intermetallic compounds, nickel aluminide Ni3Al is one of the most widely studied materials and many techniques have been developed to produce the Ni3Al-based composites (Stoloff and Alman, 1990). What makes this material unique is its anomalous thermal hardening behaviour. Depending on composition, the strength of Ni3Al increases with temperature up to approximately 600–900 ◦C. Bose and German may be among the first to investigate the feasibility of fabricating Ni3Al-based composite by MIM (Bose and German, 1988a,b; German et al., 1990; German and Bose, 1989a; Bose et al., 1992). Both prealloyed and elemental powders were used in the experiment. The preferential alignment of the fibers in the matrix was attained by a modified form of MIM, where the feedstock was extruded through a special nozzle. After debinding, the consolidation was attained by reactive sintering of elemental powders under an imposed external stress via hot isostatic compaction (HIP). Full densification was possible by appropriate selection of particle sizes, composition, thermal procedure, and pressure. As evidenced by the investigation, it is of general applicability to use relatively inexpensive elemental powder to fabricate hard-forming compounds Alman and Stoloff further explored the adaptability of MIM for production of other intermetallic matrix (NiAl, MoSi2 and TaTiAl2) composites (Alman and Stoloff, 1990, 1991a,b,c,d, 1994, 1995; Stoloff and Alman, 1991; Alman et al., 1991). The Al2O3 fibres were chopped and dispersed into the powders by mixing in alcohol. MIM was used to form the shape and the consolidation of the composites was conducted by reactive synthesis followed by hot isostatic pressing. Results indicated superior alignment was achieved with small spherical powders after injection molding. Microhardness tests revealed that the Al2O3 fibers would strengthen the substrate, especially when aligned by MIM (Alman and Stoloff, 1991d). MoSi2 is another attractive high-temperature structural material due to its high-melting point (2030 ◦C), excellent oxidation resistance and low density. Its extreme brittleness at temperature below 1000 ◦C and the low creep resistance, however, hamper the manufacture and application of this material. Adding 20 vol.% of chopped Al2O3 fibres to the MoSi2 PIM feedstock produced a MoSi2–Al2O3 composite with much improved fracture resistance and hardness (Stoloff and Alman, 1991; Alman et al., 1992). A distinct advantage of PIM in fabricating short-fibre reinforced composites is the capability of controlling the orientation of the fibres. Research results show that the expanding flow of the feedstock tend to line the fibres perpendicular to the flow direction, while the contracting flow will produce a fibre alignment parallel to the flow direction, as schematically shown in Fig. 2 (German and Bose, 1989b). By properly controlling the flow of the feedstock during injection molding, the mechanical properties along a certain direction can thus be tailored according to performance requirements. The key to successful alignment has been confirmed to be the size of the starting powder while the morphology of the powders has little effect on the degree of alignment (Alman and Stoloff, 1991a). However, this method suffers from sev-