Abstract

A primary challenge associated with TiAl alloys is their low ductility at room temperature. One approach to overcome this flaw is attaining ultrafine grains in the alloy’s final microstructure. The powder metallurgy (PM) processing route favours the synthesising of ultrafine grains in TiAl alloys. This paper features the mechanical alloying (MA) process and rapid consolidation through the spark plasma sintering (SPS) technique, which comprises the PM process. Furthermore, a second approach discussed covers microalloying TiAl alloys. An evaluation of the influence of high oxygen content is also presented, including the formation of -Al2O3. A section of the review delves into the dynamic recrystallisation mechanisms involved in elevated temperature deformation of TiAl alloys. The final section highlights the efficacy of ternary element additions to TiAl alloys against oxidation.

1. Introduction

Intermetallics are described as an ordered alloy phase formed between two metallic elements. An alloy is said to be ordered provided two or more sublattices are needed to describe its atomic structure [1]. Titanium aluminides (TiAl), over the years, have been some of the most interesting and highly researched lightweight structural materials. They have been envisaged in replacing nickel-based superalloys (NBSAs) for certain stress and temperature application ranges. The alloy design concept is quite similar to that of NBSAs. Titanium aluminides exhibit an ordered (Ti3Al) and (TiAl) phases, analogous to that of NBSA’s L12 ordered phase. Additionally, TiAl alloys exhibit a combination of enviable properties such as low density (3.9–4.2 g·cm−3), high melting point of 1733 K, high strength to weight ratio (up to 1000 MPa) that can be retained at temperatures up to 973 K, high elastic moduli, low diffusion coefficient, good creep properties up to 1173 K, and substantial resistance to corrosion and oxidation. Additionally, compared to ceramics, they exhibit good thermal conductivities and are ductile at their service temperatures while also maintaining good structural stability [2, 3].

2. Applications

Owing to the weight-saving property, inter alia other properties of titanium aluminides, it has found significant use primarily in the aerospace sector [4, 5]. So far, it has successfully been used for components in gas turbine engines where it is predicted to save weight to about 20% when used for small parts (blades, vanes, and shrouds) and about 49% when used in casings. Major turbine engine manufacturers in the industry have recognised the potential of TiAl alloys. General Electric (GE) currently uses TiAl alloys, specifically the second-generation Ti-48Al-2Cr-2Nb (at.%), for low-pressure turbine (LPT) blades in the GEnx™ used by Boeing’s 787s and 747-8s aeroplanes as shown in Figure 1. Also, SNECMA uses the same second-generation alloy for their LEAP™ engine LPT blades [6]. Pratt and Whitney and Volvo have subjected TiAl-based alloys to several rigorous tests and through their demonstrations have shown their capabilities to be used for shrouds, blade retainers, and turbine dampers [6]. The automotive industry is another sector where TiAl alloys have received increasing attention. It is used for turbine wheels of turbochargers for diesel or gasoline engines in sports cars and passenger vehicles as depicted in Figure 2 (such as Mitsubishi motors Lancer 6 model) [7]. Also, it is used for exhaust valves for high-performance racing cars [8].

3. Phases and Microstructure Evolutions

According to the phase diagram (Figure 3), TiAl alloys on solidifying goes through the single-phase region of the solid solution and, on further cooling, decomposes according to the reactions, or . Depending on the cooling rate, different phase transformations can occur, and a variety of microstructures can be attained. This gives the opportunity of controlling the microstructures to gain specific properties for peculiar applications.

The -Ti3Al phase transforms from a disordered to an ordered phase in the range of 22 to 39 at.% Al at around 1453 K. The ordered phase has a hexagonal crystal structure (DO19) with lattice parameters, a = 0.5782 nm and c = 0.4629 nm, as shown in Figure 4(a). It is known to exhibit a good high-temperature strength but with very low ductility. Additionally, it has a high rate of oxygen and hydrogen absorption, which at high temperature results in embrittlement of the alloy [9].

From 49 to 66 at.% Al, the -TiAl phase is stable and evaluated as ordered up to its melting temperature [3]. The phase has an ordered face-centred tetragonal structure (L10) with lattice a = 0.4005 nm and c = 0.4070 nm. It exhibits a tetragonality (c/a) of 1.02 (Figure 4(b)) at the equiatomic position, and this appreciates to 1.03 at increasing Al proportions [1]. It has an extremely low ductility at room temperature owing to the limited number of slip systems [10].

On the other hand, it exhibits excellent resistance to oxidation. Furthermore, the covalent bonds of atoms in the -TiAl phase translate to high specific strength and modulus associated with the phase [12]. Compared to the -Ti3Al phase, the -TiAl phase is lighter and can be used at higher temperatures.

Alloys with duplex phases, more specifically , are preferred for structural engineering applications. This is due to the flexibility in controlling the proportions of formed and phases, together with their distribution [13]. Grain morphology greatly varies depending on composition, temperature and time, cooling rate, and stabilisation temperature and time [13].

As shown in Figure 5, there are four different classifications of microstructures associated with the dual-phase TiAl alloys. They are near gamma (NG), near lamellar (NL), fully lamellar (FL), and the duplex microstructures (DM) [14].

The fully lamellar (FL) microstructure forms when fabrication temperature or heat treatment is carried out at a temperature within the pure -phase field and cooled to room temperature. During cooling, formed -Ti precipitates into alternating layers of (orders to ) and with a lamellar morphology. The grains of the microstructure are coarse and range from 200 to 1000 μm [15].

The duplex microstructure (DM) occurs when fabrication temperature or heat treatment is carried out at a temperature within the -phase field. Equiaxed grains and fine lamellar colonies of alternating plates of - and -phase characterise the duplex microstructure [17].

The near lamellar (NL) microstructure forms within the -phase field close to the -transus at temperatures between the duplex and fully lamellar. It consists of some equiaxed grains and mainly lamellar grains [11].

The near gamma (NG) microstructure is obtained in the -phase region below the eutectoid temperature. The microstructure is made up of equiaxed grains with precipitates at the grain boundaries and triple points. The microstructure is characterised by mean grain size, usually between 30 μm and 50 μm [15, 18].

4. Alloy Development

The major challenge associated with TiAl alloys is its low ductility at ambient temperatures. Over the years to date, a lot of efforts have and are channelled in improving this mechanical property. Some include (i) microalloying to form ternary or quaternary alloys (Section 4); (ii) grain refinement through the powder metallurgy processing route or by rapid solidification (Sections 5 and 6) [1921]; (iii) alteration of the morphology, volume fraction, and distribution of constituent phases by heat treatment [19, 22, 23].

4.1. Microalloying of TiAl Alloys

There have been several composition ranges studied in an attempt to improve ductility and/or a combination of other properties in TiAl alloys. The second-generation compositions of TiAl alloys in at.% are represented as follows:(i)(45–52)Ti-(45–48)Al-(1–3)X-(2–5)Y-(<1)Z where X = (Cr, Mn, and V), Y = (Nb, Ta, W, and Mo), and Z = (B, C, and Y) [15, 24]

The third-generation composition includes the following:(i)Ti-(42–48)Al-(0–10)X-(0–3)Y-(0–1)Z-(0–0.5RE), where X = (Cr, Mn, Nb, and Ta), Y = (Mo, W, Hf, and Zr); Z = (C, B, and Si). RE refers to rare earth elements [13, 16].

4.1.1. “X”-Elements Additions

The “X” elements are usually added to improve the mechanical properties of TiAl alloys through solid solution strengthening and/or precipitation hardening and also enhance ductility. Some include Nb, Cr, Mn, and V.

Nb is one element of interest present in all the compositions ranges. It has proven to enhance high-temperature strength and room temperature ductility [25, 26] but with decreased amounts of Al by shifting the -phase boundary to the left [27]. As a -stabilising element, Nb alters the -phase region of the Ti-Al system and significantly increases the eutectoid temperature [16]. These findings in recent years have significantly influenced academic dialogues on a new class of TiAl alloys termed “TNB” [3, 2832] and “TNM” [23, 3336] alloys. The concentration of Nb and/or Mo in these alloys is usually 10 at.% and includes small additions of B and C. In parallel, the addition of Nb and other BCC metals tends to modify the crystal structure of TiAl alloys to a more symmetric cubic lattice to attain the minimum number of slip systems required for plasticity, thus increasing ductility [37].

The addition of Cr and Mn has similar influence on TiAl alloys. Both elements lead to chemical strengthening (stacking fault interactions) owing to their preferential solubility in the HCP structure of TiAl alloys. As their concentration increases in TiAl alloys, they decrease the stacking fault energy and thereby further increase the separation of partial dislocations, resulting in improved ductility at elevated temperature [15, 38]. Additionally, Cr and Mn contents have been found to enhance room temperature ductility by altering the innate microstructure of TiAl-based alloys which are made apparent in decreasing the -TiAl phase tetragonality ratio (c/a) and unit-cell volume [39, 40], weakening the covalent directional Alp–Tid bonds and reducing the volume fraction of the -Ti3Al phase [41].

4.1.2. “Y”-Elements Additions

Elements such as Zr, Mo, Ta, W, and Hf form the “Y” elements. These elements, because of their high-temperature refractory properties, hinder the diffusion processes in TiAl alloys by increasing the activation energy of diffusion and subsequently decreasing the dislocations climb rate. This enhances creep and thermal stability and promotes high-temperature performance [42]. Also, Hf and W improve the creep resistance of TiAl alloys with lamellar microstructures by segregating inside the lath close to the interface and hence stabilise the interface [43]. These elements are also known to stabilise the -phase. The -phase in TiAl alloys is regarded as a ductilising constituent. The -phase has a BCC lattice which at elevated temperatures provides the adequate number of independent slip systems needed for ductility [44]. From weight and cost considerations, Ta, W, and Hf are less desired compared to Zr and Mo.

4.1.3. “Z”-Element Additions

The “Z” elements comprise low-density interstitial elements. They include C, B, Si, Y, and N. These elements form carbide, boride, silicide, and nitride precipitates. Carbon is a strong -stabiliser and is known to lead to a change in phase evolution [45]. The addition of C improves the strengthening mechanisms of TiAl alloys. It has been shown to harden effectively both the - and -phases leading to an increase in creep resistance and high-temperature strength [46]. Furthermore, C refines the lamellar microstructure of TiAl alloys by decreasing the lamellar spacing within the colonies formed upon cooling from the single -phase region. The refinement of the lamellar microstructure is premised on the segregation of C atoms to initial grain boundaries culminating in stacking fault energy reduction and therefore increases the frequency of fault formation [16]. Since the faults enhance the heterogeneous nucleation of -TiAl at grain boundaries, C effectively increases the nucleation rate of -TiAl, thus resulting in a fine lamellar structure [16]. The addition of B and Y to TiAl alloys serves as grain refining elements [24, 47, 48]. In good agreement with the Hall–Petch relationship, grain-refined microstructures enhance the mechanical properties of TiAl alloy [24]. Cheng et al. [49] asserted that increasing contents of interstitial B increases the strength of the duplex microstructure by impeding on dislocation through solute locking. The lamellar microstructure exhibits superior creep resistance because the aligned -lamellae are strong barriers to dislocation mobility. Also, dislocation generated in the -phase accumulate at -interfaces [50]. In the case of Si, its influence on TiAl alloys has been contradicting. For example, Si form silicides (-Ti5Si3, D88 structure) either by the eutectic reaction of (primary silicides) or by a eutectoid reaction of (secondary silicides) [23, 51]. These reactions depend on Si content and the alloy composition of the TiAl alloy in order to occur. These silicides enhance creep resistance by reducing dislocation mobility and also by stabilising the TiAl alloy microstructure against thermal degradation [52]. Conversely, alloying with Si has been suggested to produce extremely fine precipitates which act as weak barriers during creep. In conclusion, the presence of accumulated Si at the lamellar interfaces is argued to promote the introduction of vacancies and thus leads to destabilisation of the lamellar microstructure [53].

4.2. Presence of Oxygen

The microstructure of TiAl alloys is very sensitive to alloy compositions. The presence of interstitial elements, such as O, regarded as -stabilising elements, are known to dissolve preferentially in the -phase. Their increasing content tends to expand this phase field, raise the eutectoid temperature [54], and shift the peritectic reaction to higher Al content [55] as typified by the quasi-diagram in Figure 6.

During cooling, oxygen favours the chemical ordering of as a result of the significant difference in solubility of oxygen between the and , with the equiaxed -phase unaltered [56]. This stems from the fact that interstitial elements, especially oxygen, search for convenient sites within the L10 structure of the -phase and the DO19 structure of the -phase for optimum atomic neighbouring. Menand et al. [57] elaborated further that oxygen prefers the -phase because of the octahedral sites of the DO19 structure. These octahedral sites are made up of two kinds of cavities; the Al2Ti4 type surrounded by two atoms of Al and four Ti atoms and the Ti6 type with the cavities surrounded by six Ti atoms. The latter tend to be suitable for interstitial elements, in this case, for oxygen. Thus, the presence of high oxygen in the TiAl alloys influences diffusion, chemistry, and thermodynamic processes taking place in the alloy’s microstructure formation. A higher oxygen content reduces ductility in TiAl alloys [50, 58]. However, the presence of the oxygen atoms tends to increasingly impede on the mobility of the ordinary dislocation resulting in strength improvement of TiAl alloys [50].

4.3. -Al2O3 Formation

A likely source of oxygen addition in the processing of TiAl alloys is from their content in the starting Ti powder [59]. Additionally, the high-energy mechanical milling process, from other studies reported [60], leads to a significant increase in oxygen content which subsequently may result in an in situ reaction leading to the formation of -Al2O3 particles [61]. In support, Rahmel et al. [62] reported that Al diffusion is faster than that of Ti in -TiAl alloys, since the ratio of diffusion coefficients D(Al)/D(Ti) = 2.7 [62]. Thus, the formed -Al2O3 are promoted by high activities of Al and O in the TiAl alloy system. The formation of -Al2O3 phase with the -TiAl and -Ti3Al phases is probable as reinforced by the phase diagram of Figure 7.

Along similar lines, the saturation of oxygen in the -phase can result in the precipitation of -Al2O3. Kawabata et al. [63] asserted that when the ratio of Ti/Al is <1 or Ti/Al ratio 1 with high oxygen content, the amount of oxygen present in the -phase exceeds the solubility limit. After ordering to , the oxygen content remains higher than the solubility limit. Thus, precipitation of -Al2O3 occurs in the -phase, in , and then in the -phase. Furthermore, in other instances where the ratio of Ti/Al >1 with medium oxygen content, oxygen atoms tend to form a solid solution in the - and -phases rather than precipitation. Solid solution of oxygen in the -phase may stabilise the -phase, which may suppress the transformation. Nevertheless, retardation of this transformation causes remaining -phase to transform to by a short distance diffusion process [64].

The formation of -Al2O3 during mechanical alloying has been reported to increase the efficiency of the process [65]. Also, the presence of -Al2O3 in TiAl alloys, as suggested by Lapin et al. [50], contributes to the strengthening of the alloys.

5. Mechanical Alloying (MA)

The powder metallurgy (PM) processing route favours the synthesis of ultrafine, submicron, or nanograined microstructures and can be used in fabricating near-net-shape TiAl-based alloy products. The first stage of the PM process, wherein elemental metallic powders are blended or mixed, increases the tendency of attaining a homogeneous mixture before sintering. The process is achieved by mechanical means referred to as mechanical alloying (MA) or milling (MM). Mechanical alloying (MA) is defined as a solid-state powder processing technique involving essentially, repeated cold welding, fracturing, and rewelding of powders using high-energy ball mill equipment [66]. In the MA process, powders are plastically deformed under high-energy collision between balls and between balls and the wall of the container. As illustrated in Figure 8, the new surfaces formed during the collision promote cold welding of the powders, which are subsequently deformed plastically into a composite layered structure and eventually into new particles of different compositions. The laminate structure formation facilitates the dissolution of solute elements. According to Lü and Lai [10], the heat generated owing to the collision, lattice defects, and short diffusion distance promotes the dissolution of solute elements and formation of areas of solid solution throughout the powder matrix. Additionally, the deformation creates a high defect density, decreases the diffusion distance, and increases the powder temperature, which in the end enhances the diffusion rate. The powders undergo further fracturing and cold welding until very fine fully alloyed powders are attained with refined internal structures. For a successful MA, the right balance between the cold welding and fracturing processes is paramount [67].

The effect of mechanical milling and its process parameters in producing TiAl alloys have been studied [60, 6872].

The state of the powders (milled or unmilled) prior to sintering affects the final microstructure of TiAl alloys [73, 74]. Mechanical milling increases the tendency of a homogeneous mixture of the - and -phases owing to the high defect densities induced. These defect densities promote the nucleation of the - and -phases. The milled powders produce ultrafine grains, and their densification takes place by grain boundary sliding aided by grain or surface boundaries diffusion. On the other hand, conventional plastic deformation by gliding and twinning is active for the coarse-grained unmilled powders [73]. The atmosphere under which mechanical alloying is carried out influences the production yield of TiAl powders. For instance, the introduction of nitrogen gas after milling in argon for a certain period, known as the nitrogen shock method, further assists the fracturing of the powders by forming brittle Al supersaturated -phase solid-solution with nitrogen. The easy breakage of the formed brittle powders into fine particles results in an increase in the production yield [58, 75]. Furthermore, the tendency of attaining a homogeneous distribution of and grains in sintered TiAl-based alloys has also been shown to be favoured by an increase in rotation speed or milling time [74]. By increasing rotation speed or milling time, severe plastic deformation is likely to occur, leading to the buildup of large amounts of defect densities which favour phase transformation in the TiAl powder [74]. In addition, grinding ball sizes and the energy transferred by their impact affect solid-state reactions occurring during mechanical milling of TiAl-based alloys [76, 77]. The use of balls with different diameters in mechanical milling translates to high collision energy [10], which subsequently influences the structure of the MAed TiAl alloy. However, increasing the number of balls impacts negatively the performance of the milling process. Also, increasing the number of grinding balls raises the degree of filling of the mill causing low mobility in the balls and consequent reduction of kinetic energy transfer and milling efficiency [78]. An essential practical problem associated with the milling of TiAl alloys as well as other materials is the strong tendency of the powders sticking to the grinding media. Prolonged milling and high milling intensities presumably happen to increase the propensity of sticking due to the heating from impacts and friction, which makes the powder more ductile [60]. As a result, the severe sticking often culminates in poor microstructural homogeneity of the powders. Low milling intensity, however, slows the process of alloying [67, 77]. Process control agents (PCAs) are mostly added to help bring balance between fracture and cold welding of powders during milling, to suppress sticking. The introduction of PCAs or surfactants, which are mostly organic compounds, modifies the surfaces of deforming particles and minimises excessive cold welding. Depending on the type of PCA and milling conditions, a fraction of the PCA decomposes during milling and can contaminate the milled powders. Thus, low quantities are usually advised. Nonetheless, powder mixtures containing PCA can be pretreated by combustion synthesis (e.g., in a tube furnace) to extract the PCA before reaction occurs [79, 80]. The PCAs most commonly used in the milling of TiAl powders are benzene, stearic acid, and methanol [60, 80].

There are several studies on the constitution and structure of mechanically alloyed Ti-Al alloys. The formation of metastable FCC crystalline structures [10, 76, 81] and the formation of an amorphous phase [8285] have all been reported.

6. Spark Plasma Sintering (SPS)

Spark plasma sintering (SPS) is a rapid densification technique wherein powders are consolidated through the simultaneous application of electric current or electric field and uniaxial pressure as depicted in Figure 9. Figure 9 outlines the principal components of a typical SPS system. It is made up of a uniaxial pressing device mainly the electrodes or punches, a water-cooled vacuum chamber and atmosphere controls, a water-cooled reaction chamber, a pulsed DC generator, and regulatory systems working together with appropriate software installed on a connected computer [86].

Foremost, in the early stages of sintering, the initial application of pressure leads to the compaction of the loose powders. This causes an increase in the contact area between powder particles and, consequently, strain hardening of the particles as the pressure further increases. Sintering mechanisms for metal powders are mainly diffusion processes at the surface, grain boundaries, and lattice paths [87]. However, at the surfaces of metallic powders, there exist microscopic or submicroscopic oxide films which tend to hinder the intercrystalline diffusion of the particles [88]. Hence, the heat needed for the fusion of the contacting surfaces is provided by the application of electric current, as shown in Figure 10. As the electric current is applied, it produces sparks owing to the accumulation of high temperature at the contacting areas [89]. The spark discharges tend to break down the resistance of the surface oxides on the powders. This phenomenon is termed Joule or resistive heating of the particles. Consequently, the occurrence of Joule heating at the contact areas of the particles leads to the exposure of a fresh metal surface of higher energy state relative to the internal energy of the particles [89]. There is a reduction in surface energy as the surface crystals of the contacting surfaces coalesce as in diffusion bonding and form conductive links. However, still in this plastically deformable state, the powder compacts take the shape of the mould while mechanical pressure is simultaneously applied [88].

To better comprehend the microstructure formations and associated properties of TiAl-based alloys, the densification mechanisms and kinetics involved in the SPS process have been studied at both macroscopic and microscopic levels. Of scrutiny in these studies has to do with one of the parameters essential to the SPS process. The next paragraph looks at some examples.

Lagos and Agote [90] studied the influence of different applied loads in the densification process of the 48-2-2 alloy using elemental powders. From the study, although a lower pressure was applied to avoid the segregation of Al, maximum densification of the alloy was attained at higher temperatures up to 1573 K. Conversely, Wang et al. [91] reported that the application of higher pressure during SPS culminated in the imposition of larger sintering stresses at the contact areas of the powders. The large stresses led to the severe deformation and promoted mass transport of the particles resulting in good compaction of the TiAl alloy.

The applied current pulses and heating rate in the densification mechanisms of TiAl powders are deemed to affect the final microstructure of the alloy. High heating rates can result in near full densities of TiAl alloys, in a relatively short sintering time, which also lead to improved efficiency of the SPS process [92]. Furthermore, Zhang and coworkers [93] inferred that a high heating rate encourages grain refinement in TiAl alloys. The application of current pulses is well known from the literature to enhance mass transport through electron migration, generation of point defects, and improved defect mobility [94, 95]. At a microscopic scale with focus on the necks of the powder-powder interfaces, the applied current pulses have been shown besides being responsible for the Joule heating at the contact areas of the powders to culminate also in the particles indenting each other during densification [96]. The areas of contact are thus characterised by a recovery-recrystallisation mechanism, as depicted in Figure 11.

Various investigations on the role of holding temperatures on the densification and grain sizes of the final microstructures of TiAl powders indicated that increasing the sintering temperature during the SPS process results in the strengthening of interparticle bonding and subsequent fabrication of denser alloys [92, 9799]. Jabbar et al. [100] by using interrupted tests during the SPS process studied the effects of different sintering temperatures on the microstructural evolution of an equilibrated G4 powder alloy (Ti–47Al–1W–1Re–0.2Si at.%). The authors reported the initial microstructure of the powders had a dendritic morphology and a predominant metastable -phase. However, their heating caused the transformation of the - to the -phase with W and Re occupying the dendritic arms. At temperatures exceeding 1148 K , precipitations rich in W and Re of the B2 phase occurred. The densification process was thus attributed solely to plastic deformation.

Nevertheless, sintering at high temperatures is sometimes not recommended since it makes retaining ultrafine or nanostructures challenging [59].

In a nutshell, it is worth noting that SPS equipment varies and therefore caution should be taken in comparing sintering temperatures since different temperature measurements are employed, and none of them provides an accurate indication of the sample temperature.

The role of dwell time during the sintering process has been investigated by Wu et al. [101] as well. The authors concluded that prolonging holding time had little influence on the relative densities of the -TiAl alloy produced. However, increased holding time in the study resulted in significant improvement in the microhardness of the alloys. One of the underlying reasons for the use of SPS is to avoid grain growth in the final microstructure of alloys; thus, longer holding times are usually not encouraged.

7. Mechanical Properties

7.1. Dynamic Recrystallisation (DRX)

As a structural material, TiAl alloys’ mechanical properties are strongly dependent on the alloy composition, chemistry, microstructure, and test conditions such as temperature and strain rates. Plastic deformation of TiAl alloys mostly occurs in the -phase compared to the -phase. Ordinary and superlattice dislocations, mechanical twinning, dynamic recovery, and recrystallisation mechanisms characterise the mode of deformation in TiAl alloys. In the c-direction of the tetragonal L10 lattice of the -phase, planes with Ti and Al atoms alternate. Therefore, ordinary dislocations with Burger vector b () with non-c-components are anticipated to glide easily without altering the Ti and Al atoms at their respective positions in the lattice planes [102]. At temperatures below the brittle-to-ductile transition temperature (BDTT), the glide of ordinary dislocations and mechanical twinning activated by the glide of Shockley dislocations along are the predominant deformation mechanisms [103]. The elongation of ordinary dislocations along their screw orientation and their anchoring at various pinning points is observed in the -phase deformed at ambient temperature [104, 105]. The presence of frictional forces hindering dislocation mobility is responsible for the dislocations’ elongation in their screw direction [104]. Additionally, the many dislocation pinning points are attributed to local chemical heterogeneities such as segregation of some interstitial oxygen atoms or Al2O3 precipitates [106].

Stacking fault energy (SFE) is also essential to deformation mechanisms of materials. SFE determines the interspacing of dissociated partial dislocations and thus influences the cross-slip properties of screw dislocations [103]. Materials with low SFE tend to demonstrate more mechanical twinning, whereby they possess an additional deformation mechanism [107]. TiAl alloys possess a much lower SFE compared to ordinary metals and alloys. Moreover, the smaller SFE of the -phase (60–90 m (J·m)−2) compared to the -phase explains the occurrence of twinning in the -phase during deformation processes [108].

Of more interest in this section of the review are the dynamic recrystallisation and recovery mechanisms underlying deformation at elevated temperatures. However, for further insight into dislocation and mechanical twinning mechanisms as well as their interaction during deformation at either room or elevated temperatures, the following references are suggested to the reader [109116].

Dynamic recrystallisation (DRX) is a softening process that can occur during high-temperature deformation of TiAl alloys. Some key features of DRX include the following [117, 118]:(1)For DRX to occur, critical values of strain (associated with a critical dislocation density) and temperature (associated with a high enough mobility of grain boundaries) must be exceeded.(2)DRX is characterised by the nucleation and growth of new grains from appropriate small regions. The newly nucleated grains often form at preexisting boundaries. The driving force for grain growth during DRX, i.e., total elastic strain energy of all dislocations in the deformed microstructure, is much higher than the driving force for normal grain growth (surface energy).(3)The occurrence of DRX and formation of new dislocation-free grains are observed as sudden stress decreases in a stress-strain curve for a constant strain rate (CSR) test. As deformation advances, the dislocation density in the newly formed grains increases and the stress in the CSR test increases again. This can occur more than once and is depicted as stress fluctuations in the curve. In a constant stress creep experiment, this phenomenon produces strain rate peaks. For such fluctuations to be observed, the recrystallised volume fraction must be large enough.(4)DRX occurs simultaneously with dynamic recovery, which does not necessarily culminate in the formation of new grains. Both processes are characterised by a specific kinetic and depending on the temperature, stress, and strain rate range of the experiment, either one of the two processes can dominate, or they can occur concurrently. Whether dynamic recrystallisation occurs or not at a given set of stress, temperature, and strain rate also depends on the material.

There are two types of DRX processes. They are continuous DRX (CDRX) and discontinuous DRX (DDRX) [118]. The DDRX takes place in materials with relatively low stacking fault energies and is associated with a bulging operating mechanism. With respect to CDRX, newly formed low-angle grain boundaries gradually transform into high-angle grain boundaries and eventually turn into recrystallised grains during straining [119]. Two primary processes govern CDRX. They are progressive new grain rotation and geometric dynamic recrystallisation, as depicted in Figure 12. The CDRX usually transpires in materials with high stacking fault energies [119].

Numerous studies have reported on DRX during hot deformation of TiAl alloys. Rudolf et al. [102] attributed the overall strain compatibility of their Ti–47Al–2Mn–2Nb +0.8 vol.% TiB2 alloy to dynamic recrystallisation and not superdislocation. The authors credited the advent of the minimum creep rate in counteracting work hardening to dynamic recrystallisation. The nucleation and growth of new grains annihilate high dislocation density regions. Dislocations can be generated within the new grains, which allow for further plastic deformation. Furthermore, the new grains may nucleate and grow in crystallographic directions, which favour dislocation plasticity. So, wherever strain compatibility is required, dynamic recrystallisation provides new grains, which can deform in directions, which are needed for homogeneous macroscopic deformation. Also, during hot deformation, microstructure restoration is ascribed mainly to dynamic recrystallisation and continuous recrystallisation mechanisms commencing even at lower strain rates [121]. DRX behaviour closely relates to microstructures, and microstructures influence the mechanical properties of TiAl alloys. DRX has been reported to refine the microstructures of TiAl alloys [34]. Grain refinement over the years is deemed an effective means in improving the performance of TiAl alloys. Additionally, the formation of certain phases such as the -phase, silicide, and h-type carbides enhances DRX behaviour [34]. Al contents in TiAl alloys, on the other hand, affect DRX kinetics. Imayev et al. [122] indicated that TiAl alloys with nearly stoichiometric compositions exhibit a fast DRX kinetics while recrystallisation kinetics decreases for lower or higher Al contents. Consistent with findings by Cheng et al. [123], DRX kinetics in Al-lean TNB alloys are also slow. The occurrence of DRX in the boundaries of lamellar colonies and also in the boundaries of dynamically recrystallised grains during uniaxial hot compression tests under various deformation conditions of a TNM alloy have been reported by Xiang and colleagues [124]. The increasing DRX processes during deformation led not only to the decomposition of the lamellar structure but also to the precipitation of completely dynamically crystallised equiaxed -phase. The -phase preferentially nucleated at the boundaries of the remnant laths and the laths. Again, several research studies have hinted on the preferential nucleation of DRX at the triple colony boundaries of the lamellar structure [125127]. This is premised on deformation being the largest at the triple boundaries. Figure 13 summarises the DRX process in the lamellar colonies. Hao et al. [128] elaborated from their study with respect to Figure 13 that at the early stage of deformation, nucleation and annihilation of numerous nanotwins at adjacent lamellar interfaces occur. Concurrently, large amounts of dislocations generate from the lamellas, as depicted in Figure 13(b). With increase in strain, plenty of dislocations pile up at the twin boundaries and lamellar interfaces, as shown in Figure 13(c). As deformation continues, many dislocations are absorbed, leading to the formation of subgrain boundaries. The lamellas are then divided into many subgrains, as shown in Figure 13(d). With increased misorientation between adjacent subgrains, the subgrain boundaries transform into refined grain boundaries and eventually accomplish dynamic recrystallisation, as illustrated in Figure 13(e).

7.1.1. DRX Models

The dynamic material model (DMM) technique proposed by Prasad et al. [129], over the years, has been used to analyse quantitatively hot deformation of materials. A good example is the use of processing maps. Processing maps are contour plots of the variation of efficiency values against deformation temperature and strain rate under a given strain [130]. The efficacy of processing maps to attain optimum processing domains of metals and alloys and serve as guidance for practical plastic forming have been reported [44, 131135]. The reader is referred to [136140] for information on the fundamentals and details of processing maps. DRX occurrence has been correlated to the stable domains in processing maps [141, 142]. Essential to microstructural evolution, the event of DRX is considered the main power dissipation mechanism during hot deformation. DRX behaviour is beneficial not only to refining grain size, enhancing structural homogeneity, or improving production mechanical properties from processes such as forging but also is an efficient mechanism to attain work hardening rate and flow stress at low levels [130].

A mathematical model based on the Avrami equation has also been used to investigate DRX kinetics of TiAl alloys during hot compression testing [143, 144]. As well known, DRX of a material will start when the critical strain is exceeded. The degree of DRX subject to different processing parameters is described by its volume fraction [143]. From the theory of DRX dynamics, the relationship between DRX volume fraction and strain () can be expressed by the Avrami equation as [143, 145]where is the volume fraction of DRX grains, k is a material constant, and are critical and peak strain, respectively, and n is the Avrami constant.

From equating to flow stress, we getwhere denotes steady-state stress; and by using regression analysis, the average k and n values for the DRX kinetic model can be determined from equations (1) and (2) [143].

In summary, from the plot of vs involving the DRX kinetic model at various deformation temperatures and strain rates, the influence of hot deformation parameters on DRX behaviour and volume fraction can be assessed. Thus, an increase in strain rate and decrease in deformation temperature result in a decrease in DRX volume fraction and grain size. This is in good agreement with other reported studies [127, 146, 147]. In general, the occurrence of DRX is mainly due to dislocation glide or dislocation climb. Deformation at high temperature enhances the diffusion of atoms, cross-slip dislocation, and migration of grain boundaries, which is beneficial to nucleation and nucleus growth of DRX [148].

Other models used to analyse DRX behaviours during hot deformation of TiAl alloys are the constitutive models. Some include the Arrhenius-type constitutive model [119, 142, 149] and experimental models such as Sellars–McTegart and Hensel–Spittel models [121].

8. Oxidation Resistance

8.1. Overview and Formation of Oxides

The oxidation of metal () in pure oxygen forms an oxide () according to the reaction:

However, in applications, the presence of other parameters from the service environment complicates the oxidation process. The formation of surface oxides separates the metal surface from the environment. The formed surface oxides serve as a barrier layer which retards further oxidation reaction and decreases metal consumption. An underlying feature desired for high-temperature materials is their capability to form protective surface oxides able to resist service conditions such as hot gases and offer long-term protection during use.

The formation of surface oxides on alloys, in general, can be fundamentally described in four stages, as illustrated in Figure 14. They include, foremost, oxygen adsorption at the surface, nucleation of the oxides, and lateral growth of the nuclei, which subsequently forms a compact oxide scale [150]. For a beneficial effect, the formed surface oxides must be thermodynamically stable in the service environment, must adhere to the alloy surface, must be thermo-mechanically compatible with the metal, exhibit a low growth rate and capable of self-healing [151].

Figure 15 shows the oxidation zone at the surface of titanium-based alloys with varying Al content exposed to identical oxidation conditions. Although oxide formation can be complex, the figure provides insights into the principal differences in oxide arrangements.

A well-noticed advantage of -TiAl-based alloys during service is their resistance to oxidation in temperature ranges between 823 and 1123 K. The corrosion reaction products formed on the surface of these alloys are influenced by the alloy chemistry, exposure temperature, and duration, and the oxidising atmosphere. The primary oxides associated with TiAl-based alloys as depicted in Figure 15 include rutile (TiO2), the low-oxidation temperature corundum phase of alumina (-Al2O3), and a heterogeneous mixture of the two oxides (TiO2 + Al2O3). Despite the competition between Al-O and Ti-O bonds at the adsorbed surface of TiAl-based alloys, the high tendency of formation of TiO2 is due to the preference of oxygen atoms to Ti-rich areas [151]. As a continuous and porous TiO2 is formed at the outermost surface, Al becomes locally enriched underneath the TiO2 leading to Al-O bonds growing stronger relative to Ti-O bonds promoting the formation of Al2O3 oxides [153]. Also, this accounts for the mixed oxide zone of TiO2 and Al2O3 at the subsurface [154].

Furthermore, TiO2 is built on the oxide outermost parts rather than Al2O3 because of the fast-growing kinetics of TiO2 compared to Al2O3 [155, 156]. Therefore, due to the slower diffusion of Al, a rich Al2O3 layer is generated under the TiO2. TiO2 is characterised as less dense and off-stoichiometric n-type oxide. It is flawed with oxygen vacancies which tend to be paths for fast diffusion of oxygen, making them unprotective and significantly limiting the oxidation resistance of TiAl-based alloys [155, 156].

From Figure 15, the addition of Al decreases the width of the oxygen-affected zone. The incorporation of more Al results in a reduction of the oxide thickness, which means an improvement in oxidation resistance of the oxide [152, 157]. The Al2O3 formation is greatly desired because of its inherent protective nature as a surface oxide. This is because of attributes exhibited such as excellent compactness, good adhesion, and low growth rate [157]. TiO2 and Al2O3 make up the outermost oxides formed on the surface of TiAl-based alloys due to their outward growing nature while a mixture of the two oxides (TiO2 + Al2O3) grows inward forming inner oxides [158].

Furthermore, the influence of alloy microstructures on the formation of surface oxides has been contradicting. For instance, in a reported comparison study of near- alloys with lamellar and bimodal microstructures, improved oxidation resistance was observed from the alloy with the lamellar structure [152]. Conversely, Yamaguchi et al. [11] suggested that the oxidation resistance of TiAl alloys is independent of the microstructure of the compositions.

8.2. Influence of Ternary Elements on Oxidation Resistance

Again, one element that has received wide attention in enhancing the oxidation performance of TiAl alloys is Nb [159164]. The efficacy of Nb in improving oxidation performance is ascribed to the doping effect concept. The vacancies in the rutile lead to internal oxidation by promoting the inward diffusion of oxygen. Hence, the addition of a dopant elements such as Nb with a valence electron (+5) higher than that of Ti (+4) culminates in the decrease of defect concentration (or oxygen vacancy concentration) due to attaining electroneutrality in the oxide. This subsequently promotes the activities of Al relative to Ti, slowing the growth of TiO2 and thus favouring the establishment of a denser, continuous, and adherent Al2O3 layer which acts as a protective barrier [165]. Furthermore, the development of TiN and Ti2AlN oxides at the oxide/substrate interface serves as an excellent diffusion barrier preventing the diffusion of oxygen ions into the substrate. The presence of Nb tends to stabilise the oxide, minimise its transformation into rutile, and decrease the oxidation rate of the TiAl alloy [166].

An equal replacement of Nb is the use of Ta. Concerning their positions on the periodic table, the chemical behaviours of both elements are alike. Ta as an element is naturally known for its excellent resistance to oxidation. Additionally, as a marked advantage, Ta has a high melting temperature, thus presenting the opportunity of increasing the service temperature of TiAl alloys. Despite this knowledge, there is very limited research with contradicting findings on the influence of Ta on the oxidation resistance of TiAl alloys. Shida and Anada [167] concluded in their study that Ta additions played no role in the oxidation resistance of Ti-48.6Al (at.%) alloy. Other studies [168170] reported that Ta additions were not significantly effective compared to Nb against oxidation. In contention, studies by Popela et al. [171] on Ti-46Al-8Ta (at.%) alloy which was compared to Ti-46Al-8Nb (at.%) reported that the specific weight reduction of the Ti-46Al-8Ta (at.%) alloy was two orders lower compared to the Ti-46Al-8Nb (at.%) alloy. Furthermore, it was concluded that at temperatures of 1273 K and higher, the effect of Ta additions on oxidation resistivity was better than the Nb-containing alloy because of its ability to form a significantly denser, adherent, and protective oxides. Similarly, in a comparison study of Ti-44.8Al-6.6Ta and Ti-45.2Al-7.2Nb (at.%) investigated by Vojtěch et al. [172], the positive influence of Ta at 1273 K was attributed to the more stable and adherent oxides formed and the lower tendency of oxides spallation during cyclic oxidation. In summary, Ta tends to reduce the solubility of oxygen in TiAl alloys, therefore suppressing rutile (TiO2) formation and growth [173, 174].

The presence of Cr in TiAl during oxidation also takes the role of a doping element which restricts the permeation of oxygen and promotes preferential oxidation of Al [175]. At elevated temperatures (), Cr forms Cr2O3. Cr2O3 has a corundum structure like Al2O3 and is also a protective oxide owing to its parabolic growth mechanisms [176]. Cr in TiAl alloys has been suggested to enhance adhesion of oxide mixtures of Al2O3 and TiO2 due to its solubility in Al2O3 and compatibility with TiO2 [175]. On the other hand, the potency of Cr in positively influencing oxidation resistance of TiAl alloys is largely dependent on its concentration. Cr additions higher than 8 at.% have beneficial effects in TiAl alloys, whereas concentrations lower than 4 at.% are detrimental [177].

Also, W and Mo improve oxidation resistance by forming W- or Mo-+enriched -Ti phases such as Ti2AlMo and/or -Ti phase at the metal side of the oxide/metal interface [178]. Oxygen solubility in the formed layer is low, whereas Al diffusion is fast [178]. This drives not only the supply of Al into the oxide scale but also leads to the formation of Al2O3 at the oxide/metal interface [167].

The addition of Y in the right amount influences the oxidation resistance of TiAl alloys. The incorporation of Y significantly improves long-time oxidation resistance by retarding rutile formation, favouring Al2O3 formation and growth, and also strengthening the spallation resistance of formed scales [179]. During oxidation, Y promotes the generation of a Y-rich layer at the immediate surface of TiAl alloys. Y has a strong affinity to oxygen during high-temperature oxidation and hence forms Y2O3 oxide and (Y, Al) O-type oxides [180]. While formed Al2O3 acts as diffusion barriers impeding oxygen ingress, the inner (Y, Al) O-type oxides tend to reduce thermal stress by modifying the oxide layer’s grain structure. Hence, the residual stress between the oxide scale and substrate may be removed [181]. The beneficial effect of Y on the scale spallation of TiAl alloys is attributed to proposed mechanisms such as [181183](1)The reactive-element-oxide particles encourage mechanical keying of the oxide scale on the substrate(2)The reactive elements tend to serve as vacancy sink sites resulting in the suppression of vacancy coalescence at the alloy-oxide interface for void formation(3)The segregation of Y at the oxide grain boundaries blocks diffusion paths and thus decreases growth stress in the scale(4)The formation of a fine-grain layer structure improves the plasticity of the oxide scale

Another concept used in improving the oxidation resistance of TiAl alloys includes the incorporation of halogens (F, Cl, Br, and I) termed as the halogen effect. The addition of halogens has been investigated and concluded to support selective Al transport and improve adhesion between the formed Al2O3 scale and the metal surface. For detailed information, the reader is referred to the following references [184189].

9. Conclusion

Titanium aluminide- (TiAl-) based alloys have gained immense recognition as a new group of weight-saving materials. Their properties, to some extent, better the shortcomings of conventional titanium alloys at higher temperatures.

The effectiveness of the powder metallurgy route of manufacturing TiAl alloys, in recent years, has been promising. The combined use of high-energy mechanical milling and spark plasma sintering increases the tendency of attaining ultrafine and homogeneous microstructures at a cheaper cost. The PM route, together with microalloying, enhances specific properties despite the tradeoff of others. The addition of niobium (Nb) has stood out compared to other elements in achieving the right balance of properties at both ambient and elevated temperatures.

In summary, the competent usage of TiAl-based alloys as structural engineering materials rests on the application of the most effective tailored microstructures attained through controlled processing and appropriate alloy modifications that can promote and stabilise new microstructures. Thus, there is the need to understand the fundamentals of alloy design, processing, characterisation of the microstructures, microstructure and property relationship, behaviour or performance of these new alloys as well as the influence of the environment on these properties during service.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.