Influence of Microstructure and Texture on Tensile Properties of an As-Rolled Ti₂AlNb-Based Alloy

Abstract

Ti₂AlNb-based alloys are widely used in aerospace applications due to their excellent high-temperature mechanical properties. This study aims to investigate the texture, microstructural evolution, and phase transformation behavior of Ti₂AlNb-based alloy sheets during heat treatment and their effects on tensile properties. During heat treatment, B2 → O phase transformation occurs at 550 °C and 650 °C, while Ostwald ripening takes place at 700 °C and 850 °C. The α₂ phase undergoes spheroidization around 1000 °C due to grain boundary separation and recrystallization. Additionally, the B2, O, and α₂ phases all exhibit strong textures. The B2-phase texture follows a cubic orientation ({100}<001>), rotated ~30° around the normal direction (ND). The O-phase texture consists of a strong {100}<010> rolling texture and a weaker texture component <001>//RD, influenced by the B2-phase texture, rolling deformation, and variant selection during O-phase precipitation. Each B2 grain generates four variants, forming distinct O-phase textures within the same grain. The α₂-phase texture exhibits typical rolling textures, [0001]//TD, <$\overline{1}$2$\overline{1}$0>//ND, and {11$\overline{2}$0}<01$\overline{1}$0>, remaining stable after heat treatment. Tensile tests show that the rolled sheet has better ductility along the rolling direction (RD), while the transverse direction (TD) demonstrates higher yield strength (up to 1136 MPa). The anisotropy in tensile properties is mainly attributed to the O-phase texture, with minor contributions from the α₂-phase and B2-phase textures. These findings provide a theoretical basis for optimizing the mechanical properties of Ti₂AlNb-based alloys.

1. Introduction

In 1988, Banerjee et al. [1] first identified the orthorhombic (O) phase in the Ti-25Al-12.5Nb (at.%) alloy, expanding the application potential of the Ti₃Al-Nb alloy system [2,3]. The addition of Nb enhances the mechanical properties of alloys [4,5,6,7]. Among these, Ti₂AlNb-based alloys have attracted significant attention in the aerospace field due to their excellent high-temperature strength, fracture toughness, room-temperature ductility, oxidation resistance, and superplastic deformation capability [8,9,10,11,12]. Typically, Ti₂AlNb-based alloys consist of primary phases: the B2 phase with a body-centered cubic (BCC) structure, the α₂ phase with a hexagonal close-packed (HCP) structure, and the O phase with an orthorhombic structure. As the matrix phase, the B2 exhibits good ductility owing to its numerous slip systems. In contrast, the α₂ phase has fewer slip systems and is typically brittle, while the O phase presents intermediate properties between these two phases [13,14,15,16,17].

The mechanical properties of alloys are significantly influenced by composition, microstructural morphology, and microstructural homogeneity. Over the past few decades, researchers have successfully fabricated Ti₂AlNb-based alloys with various microstructures, such as equiaxed, bimodal, and lamellar, through thermal processing techniques including casting, isothermal forging, and hot rolling [18,19,20,21,22,23,24]. Previous studies from the research group have established that the fine lamellar O phase in as-cast Ti₂AlNb-based alloys not only provides good plasticity and strength but also enhances fatigue resistance [25]. Forged alloys with a bimodal O phase exhibit the highest yield strength and tensile strength but the lowest ductility [2]. The rolling process has proven effective in improving the mechanical properties of as-cast alloys. To address the poor room-temperature ductility of Ti₂AlNb-based alloys, Sui et al. [26] and Wang et al. [27] obtained rolled alloys with both ductility and strength via hot-rolling after spark plasma sintering (SPS) of pre-alloyed powders. Microstructural inhomogeneity may significantly affect the reliability of mechanical properties. Zavodov et al. [28] investigated the influence of deformation bands (DB) in Ti-22Al-25Nb (at.%) alloy rolled sheets on the ultimate tensile strength (UTS) in tensile tests. The results indicated that the presence of DB and its area ratio significantly reduced material strength, with single-crystal structures and unfavorable orientations being the primary contributing factors.

The texture-induced anisotropic behavior of Ti₂AlNb-based alloys plays a critical role in advancing their application in aeroengine components [29]. Zhang et al. [29] demonstrated that the yield strength anisotropy of Ti-22Al-25Nb alloy, isothermally forged in the B2-phase region, is predominantly governed by the O-phase texture. The deformation temperature and phase composition significantly influence the morphology and texture development of the O phase. Dey et al. [30] observed that during deformation in the two-phase region, the O phase tends to become equiaxed, with its texture intensifying as deformation progresses. Conversely, deformation in the single β-phase region, followed by cooling, results in the formation of coarse lamellar O phases that inherit the deformation texture of the β phase. Rollett et al. [31] investigated the as-rolled Ti-22Al-23Nb (at.%) and found that the BCC phase exhibited a strong (001)<110> texture, while the orthorhombic and hexagonal phases showed weaker textures, suggesting minimal variation in yield strength across different directions. Wang et al. [32] further validated this hypothesis, reporting tensile strengths of 1110 MPa and 1157 MPa in the rolling direction (RD) and transverse direction (TD), respectively, with elongations of 5.7% and 2.4%. Liu et al. [33] investigated the influence of α-phase texture on yield strength through theoretical calculations of Schmid factors (SFs) and critical resolved shear stress (CRSS), concluding that the α phase is responsible for the observed strength anisotropy in the alloy.

Current research has discussed the influence of rolling processes on microstructure and mechanical properties, establishing correlations between texture and mechanical performance using parameters such as SF and CRSS [25,29,30,31,33]. However, few studies have examined how subsequent heat treatment affects the microstructure and texture of rolled sheets, and a detailed explanation of the formation mechanism of O-phase texture remains lacking. This study systematically investigates the evolution of texture during heat treatment in rolled Ti₂AlNb sheets with complex microstructures. The formation mechanism of the O-phase texture is analyzed from two perspectives: the inheritance of B2-phase texture and the variant selection of the O phase. Furthermore, the impact of multiphase textures on the anisotropic tensile properties of the alloy is examined, offering novel insights into texture regulation for rolled Ti₂AlNb-based alloys.

2. Materials and Methods

2.1. Materials

The present study focused on a Ti₂AlNb-based alloy sheet with a nominal composition of Ti-22Al-23Nb (Mo, V, Si) (at.%) (Shanghai, China). The alloy sheet was produced through a series of processing steps: First, the alloy ingot was first fabricated via three cycles of vacuum arc remelting (VAR), followed by hot isostatic pressing (HIP) to homogenize the microstructure. Next, the ingot was subjected to multi-directional isothermal forging to refine the grain structure. Finally, the sheet was manufactured using pack precision hot rolling combined with intermediate heat treatment, yielding a hot-rolled sheet with a thickness of 2.5 mm.

2.2. Methods

Rectangular samples measuring 10 × 8 × 2.5 mm (Figure 1a), along with tensile specimens oriented in the rolling direction (RD) and transverse direction (TD), were cut from the fabricated hot-rolled sheet. The rectangular samples underwent heat treatment at temperatures of 550 °C, 650 °C, 700 °C, 850 °C, 900 °C, 950 °C, 1000 °C, 1150 °C, and 1200 °C, respectively. The possible phase compositions at different temperatures were referenced to the Ti-22Al-xNb (at.%) pseudo-binary phase diagram (Figure 1d). To prevent oxidation during the heat treatment process, all samples were hermetically sealed within quartz tubes. Following a 12 h heat treatment, the quartz tubes were carefully fractured to enable water quenching, thereby ensuring the preservation of the high-temperature microstructure. Schematic diagrams illustrating the samples and heat treatment procedures are presented in Figure 1a,b.

To analyze microstructural evolution and texture formation mechanisms, and to predict tensile properties during the heat treatment process from the perspectives of morphology, phase content, texture, and grain orientation, detailed characterization was conducted using scanning electron microscopy (SEM) (Carl Zeiss AG, Jena, Germany), electron backscatter diffraction (EBSD) (Oxford Instruments, Oxford, UK), X-ray diffraction (XRD) (Bruker AXS, Karlsruhe, Germany), and transmission electron microscopy (TEM) (TALOSF200XG2) (Thermo Fisher Scientific, San Diego, CA, USA). The step size for the EBSD scan was configured to 0.05 μm. It should be noted that samples for EBSD analysis were selected at 850 °C, 900 °C, 950 °C, 1000 °C, and 1150 °C, with the aim of observing the changes in the content of the O phase, the grain size of the B2 phase, and the spheroidization and dissolution of the α₂ phase.

Sample preparation involved mechanical polishing and electro-polishing. The mechanical polishing utilized Struers OP-S metallographic polishing suspension, while the electro-polishing solution consisted of 6% perchloric acid, 34% n-butanol, and 60% methanol. During electro-polishing, liquid nitrogen was used to cool the electrolyte, maintaining a temperature between −40 °C and −30 °C. The DC power supply voltage was set to a maximum of 60 V with a constant current of 0.8 A, and the polishing duration was controlled to 90 s. After polishing, the samples were thoroughly rinsed with water, followed by ethanol, and then dried. For backscattered electron (BSE) observations, the polished samples were etched using Kroll’s reagent (3% HF + 5% HNO₃ + 92% H₂O).

The room-temperature tensile properties of the Ti₂AlNb-based alloy sheet were evaluated along both the RD and TD using an Instron 5982 tensile testing machine (Instron, Norwood, MA, USA). For each orientation, three parallel specimens were prepared by sequential grinding with abrasive papers followed by mechanical polishing. All the tensile tests were conducted at a constant strain rate of 1 × 10⁻³ s⁻¹. The orientation, geometry, and dimensions of tensile test specimens are presented in Figure 1c.

3. Results

3.1. Microstructure of the Rolled Sheet and Its Evolution During Heat Treatment

3.1.1. XRD and TEM

The XRD patterns in Figure 2 reveal the presence of three distinct phases in the as-rolled sheet: B2, α₂, and O phases. The TEM image in Figure 2c reveals the presence of needle-like and blocky O phase (points 1 and 2), coarse plate-like α₂ phase (point 3), and B2 phase intercalated between the O and α₂ phases (point 4). Notably, some α₂ phase is observed to be intercalated within the needle-like O phase (point 2). Furthermore, the O phase and α₂ phase exhibit a well-defined crystallographic orientation relationship, specifically ($\overline{1}$31)_O//(01$\overline{1}$0)_α2.

A close examination of the XRD patterns reveals that, as the heat treatment temperature increases, the peaks corresponding to the O phase almost disappear in the sample heat-treated at 900 °C. In the microstructures heat-treated between 900 °C and 1000 °C, only B2 and α₂ phases remain, while a single-phase microstructure is observed at 1150 °C. Additionally, by examining the main peaks of the B2 and O phases in the as-rolled state, and after heat treatment at 550 °C, 650 °C, 700 °C, and 850 °C, it is evident that the main peak of the O phase in the as-rolled state is lower than that of the B2 phase. However, at 550 °C and 650 °C, the main peak of the O phase becomes higher than that of the B2 phase. At 700 °C, the main peak of the O phase is again slightly lower than that of the B2 phase, and by 850 °C, the main peak of the B2 phase is significantly higher than that of the O phase. These observations suggest that the B2 → O transformation occurred during heat treatment between 550 °C and 650 °C, while at 700 °C, part of the O phase dissolved back into the B2 phase, and at 850 °C, the dissolution became more pronounced.

3.1.2. SEM

The SEM images of the as-rolled and heat-treated samples are shown in Figure 3. It should be noted that in Ti₂AlNb-based alloys, the B2 matrix typically appears light in BSE images, while the α₂ phase (stoichiometric chemical formula: Ti₃Al) exhibits near-black contrast [24,35]. The O phase (stoichiometric chemical formula: Ti₂AlNb) consistently displays a gray contrast in BSE imaging [23]. In the as-rolled Ti₂AlNb sheet, the B2 phase serves as the matrix, while the α₂ phase appears as primary coarse plates and secondary equiaxed grains. The O phase is distributed in the form of fine laths, edge layers, and a small number of blocky structures. During the rolling deformation process, the plate-like α₂ phase was significantly crushed, forming grooves and bent plate-like features. The temperature range of 550–650 °C corresponds to the stable region of the O phase. After 12 h of heat treatment, the lath-like O phase coarsened and became densely distributed within the B2 phase, while the content of the B2 phase decreased. At 700 °C, which lies within the B2 + O two-phase region, the O phase coarsened relative to the as-rolled microstructure but exhibited a sparser distribution compared to samples heat-treated at 550 °C and 650 °C. This indicates that the O phase primarily coarsened due to phase transformation in the single-phase region, while in the two-phase region, it further coarsened and partially dissolved into the B2 phase. This phenomenon became more pronounced in the 850 °C heat treatment condition. The microstructural characteristics observed via SEM fully correspond well to the earlier XRD analysis. After heat treatment at 900 °C, the lath-like O phase nearly completely dissolved. It is necessary to note that due to the high Nb content in Ti₂AlNb-based alloys, the diffusion of elements is relatively slow [29]. Once the O phase nucleates, its growth kinetics remain sluggish, resulting in the O phase always being small in size throughout the rolled sheet microstructure. Therefore, high-magnification electron microscopy is required to clearly observe it. The α₂ phase remained relatively stable in both the O single-phase region and the B2/β + O two-phase region but partially dissolved upon entering the B2/β + O + α₂ three-phase region. At 1000 °C, some of the crushed and bent α₂ phases underwent significant fragmentation and exhibited spheroidization, then completely dissolved upon further heating.

3.1.3. EBSD

EBSD characterization was performed on selected heat-treated microstructures of interest (Figure 4). In the EBSD maps, brown, light blue, and yellow represent the B2 phase, α₂ phase, and O phase, respectively. In Figure 4d–f, the white lines indicate high-angle grain boundaries (HAGBs).

Table 1. Phase content of Ti₂AlNb-based alloy after heat treatment at different temperatures.

Phase	As-Rolled	850 °C	900 °C	950 °C	1000 °C	1150 °C
B2	35.8%	45.5%	70.3%	86.7%	91.3%	100%
O	41.3%	19%	2.6%	0	0	0
α2	18.7%	22.8%	22.8%	12.5%	7%	0

summarizes the phase fractions in different heat treatment conditions. The results show that as the heat treatment temperature increased, the B2-phase content gradually increased after 850 °C. Concurrently, the B2 grains exhibited progressive size enlargement when the temperature exceeded 950 °C, as shown in Figure 4f and Figure 5f,g. The grain size of B2 phase increased from approximately 14.5 μm to over 100 μm. In contrast, the O-phase content gradually decreased, with its dissolution occurring between 900 °C and 950 °C. The α₂ phase remained relatively stable below 900 °C but began to dissolve beyond this temperature, with significant dissolution of secondary phases observed at 1000 °C. The primary plate-like α₂ phase, located at grain boundaries, exhibited a spheroidization trend. No phase transformation between α₂ phase and the O phases was observed during the heat treatment process.

Figure 5 presents the statistical results of misorientation angles between adjacent grains and grain sizes for each phase in the Ti₂AlNb-based alloy. Figure 5a–e also include the proportions of low-angle grain boundaries (2–15°) and high-angle grain boundaries (>15°). In the as-rolled state, the misorientation angles between adjacent B2 grains are mainly concentrated in the low-angle range, indicating that the B2 phase was deformed during rolling, resulting in the formation of numerous substructures. As the heat treatment temperature increases, the misorientation angles between adjacent B2 grains shift towards high angles, suggesting partial recrystallization of the grains. Heat treatment at 1000 °C led to nearly complete recrystallization of the B2 phase, while at 1150 °C, the grain orientations became entirely random. The B2 grains gradually grew, but when the α₂ phase was not fully dissolved, the pinning effect of second-phase particles inhibited the growth of B2 grains. At 950 °C and 1000 °C, the average grain sizes are approximately 14.5 μm and 18.6 μm (Figure 5f,g), respectively. Once the α₂ phase was completely dissolved, the B2 grains grew rapidly. Due to the excessively large grain size, the number of complete B2 grains captured was limited. Based on metallographic, SEM, and EBSD data, it can be estimated that after heat treatment at 1150 °C, B2 grain sizes generally exceeded 500 μm, forming equiaxed large grains.

The misorientation angles between adjacent α₂ grains are concentrated near 10°, 30°, 60°, and 90° (Figure 5h). In titanium alloys, 12 variants can form within a single β grain, with each variant having an equal probability of formation. These 12 variants exhibit specific misorientation angles between each pair [36,37]. In the Ti-6Al-4V alloy, theoretical calculations predict five types of misorientation angles: 10.5°, 60°, 60.832°, 63.262°, and 90°, with a ratio of 1:2:4:2:2 in the absence of variant selection [38,39,40]. The distribution of misorientation angles in the as-rolled alloy suggests that some form of variant selection may occur during the formation of the α₂ phase, leading to the development of α₂ texture. At 950 °C, the number of high-angle α₂ grains significantly decreases, likely due to the dissolution of secondary equiaxed α₂ grains. In the sample heat-treated at 1000 °C, the misorientation angles between adjacent α₂ grains show a tendency to shift towards higher angles, indicating that some coarse plate-like α₂ grains undergo recrystallization.

3.2. Texture of the B2, O, and α₂ Phases

The pole figures of the B2 phase exhibit extensive regions of green, indicating the presence of texture (Figure 6). This texture is characterized by a cubic orientation rotated approximately 30° around the normal direction (ND). Following heat treatment at 900 °C, the texture was observed to weaken; subsequently, a {111}//rolling plane and <111>//ND texture emerged, likely forming during the recovery process of the B2 phase. The texture remains evident after heat treatment at 950 °C but is completely eliminated at 1150 °C as the second phase fully dissolves, leading to random orientations. Throughout the heat treatment process, an increase in temperature correlates with a gradual enhancement in recrystallization of the B2 phase, accompanied by a corresponding weakening of its texture.

The rolled sheet exhibits both coarse plate-like and secondary equiaxed α₂ grains, with the coarse plate-like α₂ phase demonstrating a pronounced preferred orientation. The pole figures and inverse pole figures presented in Figure 7a illustrate textures such as [0001]//TD, <$\overline{1}$2$\overline{1}$0>//ND, and {11$\overline{2}$0}<01$\overline{1}$0> rolling texture. It should be noted that the obtained texture of the α₂ phase differs from that of single-phase hcp metals such as magnesium alloys and titanium alloys after rolling deformation [41,42]. However, Ti₂AlNb is a multiphase alloy, and the α₂ phase is different from the conventional hcp phase [43]. Its deformation mechanism selection and texture evolution during rolling are strongly influenced by the B2-phase matrix, thus resulting in a texture evolution outcome distinct from that of single-hcp-phase alloys. Similar [0001]//TD oriented texture was also reported in [44,45]. The α₂ phase remained relatively stable within the B2 + O two-phase region. Upon entering the three-phase region (~940 °C), the secondary equiaxed α₂ phase dissolved first, resulting in a slight increase in texture intensity (Figure 7c,d). Heat treatment does not significantly affect the preferred orientation of the α₂ phase.

The consistent green color observed in the lath-like O phase indicates a preferential orientation along the <010> direction (Figure 7e). The O phase displays a pronounced {100}<010> rolling texture, alongside a weaker texture component <001>//RD, which correspond to the green and red regions in the inverse pole figure (IPF), respectively. As the O phase undergoes dissolution, it is evident that the intensity of this texture gradually diminishes. Furthermore, it can be noted that the texture of the O phase exhibits a correlation with both the cubic-like texture of the B2 phase and the texture of the α₂ phase.

3.3. Tensile Properties

The rolled sheet exhibits notable anisotropy in its tensile properties. Room-temperature tensile tests (Figure 8,

Table 2. Room-temperature tensile properties of three parallel samples of Ti₂AlNb-based alloy: yield strength (YS), ultimate tensile strength (UTS), and elongation (El).

	Strain Rate, s−1	YS, MPa	UTS, MPa	El%
RD-1	1 × 10−3	936	1032	3.1
RD-2	1 × 10−3	953	1063	4.0
RD-3	1 × 10−3	942	1044	2.9
TD-1	1 × 10−3	1136	1194	1.8
TD-2	1 × 10−3	1018	1150	1.5
TD-3	1 × 10−3	1055	1194	1.5

) indicate that the alloy exhibits greater ductility in the RD while achieving a high yield strength in the TD. The yield strength reaches an average value of 1070 MPa in the TD, while showing a lower average value of 944 MPa in the RD in Table 2.

4. Discussion

4.1. Spheroidization of the α₂ Phase and Coarsening of the O Phase

After transitioning from the B2/β + O two-phase region to the B2/β + O + α₂ three-phase region, the secondary equiaxed α₂ phase was observed to dissolve first. However, following heat treatment at 1000 °C, a substantial amount of equiaxed or nearly equiaxed α₂ phase remains in the microstructure, indicating that spheroidization occurred at this temperature. Both prior deformation and heat treatment temperature are critical factors influencing static spheroidization [46]. During the rolling process, the plate-like α₂ phase underwent elongation, crushing, or deformation, resulting in the formation of numerous substructures within the plates [47,48]. At elevated temperatures, the B2 phase can readily infiltrate these regions, leading to the disintegration of the plates. In this context, the spheroidization mechanism of the α₂ phase is characterized by grain boundary separation [46,49,50,51]. Additionally, the formation of numerous subgrains within the α₂ plates during deformation leads to the accumulation of stored deformation energy, which subsequently drives recrystallization during heat treatment. Therefore, recrystallization occurring during heat treatment also plays a significant role in the spheroidization of the α₂ phase [46,52]. Another mechanism that facilitates spheroidization in the Ti₂AlNb-based alloys involves the mixed growth of the O phase within elongated, plate-like α₂ phases, as illustrated in Figure 4b and case ③ of Figure 9a. During heat treatment, the O phase undergoes a transformation into the B2 phase, often occurring in the necking regions of α₂ plates. This phenomenon resembles a termination migration mechanism [53], where elements diffuse from high-curvature areas to smoother regions. Figure 9a illustrates three pathways for spheroidization. In practice, the grain boundary separation mechanism typically appears during the early stages of heat treatment, with the release of stored deformation energy at subgrain boundaries driving the transformation of the α₂ phase into the B2 phase. The α₂ grains that undergo spheroidization via grain boundary separation and element migration exhibit similar orientations. Conversely, α₂ grains undergoing spheroidization through recrystallization are characterized by high-angle grain boundaries, leading to significant misorientation. It is crucial to note that complete spheroidization still requires adjustment through the tip migration mechanism [54].

The thickness of the O-phase lamellae exhibits a strong dependence on temperature, increasing as the temperature rises. This phenomenon can be attributed to the fact that higher temperatures supply greater energy for diffusion, thereby promoting lamellar growth [55], as described by the Arrhenius equation. At temperatures of 550 °C and 650 °C, which are close to the O single-phase region, the transformation of the B2 phase into the O phase led to the coarsening of O-phase lamellae. This growth process in the lamellar O phase is governed by energy minimization and the equilibrium of atomic concentration between the lamellar O phase and the B2 matrix. A new stable state is attained by reducing the total interfacial area between O and B2 phases while preserving atomic concentration equilibrium [55,56,57]. When the temperature rises to 700 °C and 850 °C, within the B2/β + O two-phase region, some unstable and smaller O phase dissolved, while larger and more stable lamellae continue to grow, as illustrated in Figure 9b. After a heat treatment duration of 12 h, a portion of the O phase transformed into the B2 phase. Meanwhile, the remaining O phases coalesced with one another, resulting in coarsening and columnar growth of the O phase. This phenomenon can be elucidated by Ostwald ripening.

4.2. Formation Mechanism of Texture

In the rolled Ti₂AlNb-based alloys, both B2 and O phases exhibit pronounced texture characteristics, with a specific crystallographic orientation relationship existing between the two phases [36,37]. During the rolling process, the α₂ phase precipitates from the B2 phase at elevated temperatures while simultaneously undergoing deformation induced by rolling. It is noteworthy that heat treatment alone does not induce the formation of the {11$\overline{2}$0}<uvw> texture in the α₂ phase [46,47]. Therefore, the {11$\overline{2}$0}<uvw> texture represents a hot deformation texture developed during rolling. The O phase characterized by both strong and weaker texture components precipitates from the B2 phase, with textures corresponding to those of the cubic structure of the B2 phase, specifically, {100}O//{001}B2//rolling plane and <100>O//{100}B2//ND. The strongly textured O phase is designated as O1, while that with secondary strength is referred to as O2. Both O1 and O2 primarily maintain a 90°/[100] rotation relationship, as illustrated in Figure 10b,c. For this paper, the remaining O phase was not analyzed for the time being; its content is very small, being mostly edge O phase formed by the peritectic reaction between the B2 phase and the α₂ phase, which could introduce interference in the analysis.

The formation of the O-phase texture is influenced by the variant selection effect that occurs during its precipitation from the B2 phase. Under ideal conditions, a single BCC unit cell has the potential to generate 24 variants of the O phase [58]. However, due to the distinct ordered structures of both the B2 and O phases compared to typical crystal structures, only 12 variants are observed within a single BCC unit cell, with each variant manifesting with equal probability [29,59,60,61]. However, in actual production processes, due to the influences such as deformation and thermal effects, variant selection may occur, leading to a significantly higher probability of forming certain specific O-phase variants [40,58,62,63]. This ultimately results in texture developments.

The correspondence between the {110} plane and <111> direction of the B2 phase and the {001} and {110} planes of the O phase, as illustrated in Figure 10c, indicates that during precipitation, the O phase maintains a Burgers vector relationship (BOR) with the B2 phase. However, it is evident that the B2-phase grains experience deformation and rotation during rolling, resulting in non-ideal single grains. Furthermore, since a single BCC grain has six completely equivalent {110} crystal planes and four equivalent <111> crystal directions, the eight focal points of the {110} plane in Figure 10c indicate that the B2 phase with strong texture contains grains of four orientations. To analyze these orientations, this study divides the analysis area into four grain types, illustrated in Figure 10a, with pink, green, blue, and yellow denoting G1, G2, G3, and G4, respectively. Extensive pole figure analysis of the O phase reveals the variant conditions of different types of B2 grains, as summarized in

Table 3. Precipitation of O-phase variants in B2 grains with different orientations in the as-rolled Ti₂AlNb-based alloy sheets.

	O1		O2
G1	V1	($\overline{1}$10)B2//(001)O, [11$\overline{1}$]B2//[1$\overline{1}$0]O	V3	(110)B2//(001)O, [$\overline{1}$11]B2//[1$\overline{1}$0]O
	V2	($\overline{1}$10)B2//(001)O, [111]B2//[110]O	V4	(110)B2//(001)O, [1$\overline{1}$1]B2//[110]O
G2	V1	($\overline{1}$10)B2//(001)O, [111]B2//[1$\overline{1}$0]O	V3	(110)B2//(001)O, [$\overline{1}$11]B2//[110]O
	V2	($\overline{1}$10)B2//(001)O, [11$\overline{1}$]B2//[110]O	V4	(110)B2//(001)O, [1$\overline{1}$1]B2//[1$\overline{1}$0]O
G3	V1	(110)B2//(001)O, [1$\overline{1}$1]B2//[110]O	V3	($\overline{1}$10)B2//(001)O, [111]B2//[1$\overline{1}$0]O
	V2	(110)B2//(001)O, [$\overline{1}$11]B2//[1$\overline{1}$0]O	V4	($\overline{1}$10)B2//(001)O, [11$\overline{1}$]B2//[110]O
G4	V1	(110)B2//(001)O, [$\overline{1}$11]B2//[110]O	V3	($\overline{1}$10)B2//(001)O, [111]B2//[110]O
	V2	(110)B2//(001)O, [1$\overline{1}$1]B2//[1$\overline{1}$0]O	V4	($\overline{1}$10)B2//(001)O, [11$\overline{1}$]B2//[1$\overline{1}$0]O

. Figure 11 illustrates the analysis process of the variants in G1, showing that each type of grain could only precipitate four distinct variants; within a single type of grain, O1 and O2 are composed of two variants each. Although there is inconsistency in the precipitation relationship between various types of B2 grains and the O phase, this relationship is precisely what drives the formation of a strong texture in the precipitated O phase. Overall, rolling induces deformation in the B2 phase, contributing to the development of a robust texture. The O phase subsequently inherits this texture from the B2 phase during its formation. Furthermore, dislocations generated during the rolling process enhance the variant selection effect in the O phase [29,58,64], ultimately contributing to the establishment of a strong texture in this phase.

4.3. Relationship Between Texture and Tensile Properties

Since the microstructure of the rolled sheet contains a high proportion of textures, it is necessary to explain the influence of texture on its tensile properties. Although α_2-Ti₃Al, Mg, and Ti all belong to the P₆₃/mmm space group, α₂-Ti₃Al exhibits a D0₁₉ ordered hexagonal structure characterized by highly ordered atomic arrangements [65,66,67]. Twinning involves coordinated shear displacement of atoms along specific crystallographic planes; however, in an ordered structure such as D0₁₉, the fixed atomic positions resist such deformation. Twinning deformation disrupts the long-range order, resulting in the formation of high-energy anti-phase boundaries or complex defects, which significantly elevates the energy barrier for twinning [43,68,69,70,71]. Furthermore, as an intermetallic compound, α₂-Ti₃Al features a combination of metallic and covalent bonding. The strong directional nature of covalent bonds impedes shear deformation, thereby hindering the nucleation and propagation of twins [66,72]. Consequently, compared to twinning, which would destroy the ordered structure, dislocation slip is more energetically favorable and thus more likely to occur [43,69,71]. The slip behavior of each phase in the material plays an important role in determining yield strength. From the perspective of single crystals, the yield strength of a material is governed by the CRSS and the Schmid factor (SF):

$${\mathsf{\sigma }}_y={\displaystyle \frac{\mathrm{CRSS}}{\mathrm{SF}}},$$

(1)

For polycrystalline materials, it can be expressed by the yield strength (σ_i) and volume fraction (f(V_i)) of each orientation grain:

$${\mathsf{\sigma }}_y={\sum }_{\mathrm{i}=1}{\mathsf{\sigma }}_{\mathrm{i}}*\mathrm{f}({\mathrm{V}}_{\mathrm{i}}),$$

(2)

From Equation (1), it is evident that a higher SF and a lower CRSS will result in a lower yield strength. Equivalent slip systems have the same CRSS, whereas the non-equivalent slip systems differ. Based on the method introduced by Zhang et al. [60], we evaluated the yield strength in the TD and RD of the rolled sheet. Since we only focus on the influence of texture and the rolled sheet contains a high proportion of texture, the yield strength ratio between the RD and TD directions for the B2 and α₂ phases can be simply determined based on the SF. However, for the O phase, which exhibits a weaker texture component, the differences in grain orientations must also be considered when calculating the yield strength ratio.

Figure 12 and Figure 13 illustrate the distribution of Schmid factors for each phase under tensile loading in different directions, along with the corresponding inverse pole figures. Since the Schmid factors for the {123}<111> slip system in the B2 phase and the {11$\overline{2}$2}<$\overline{1}\overline{1}$23> pyramidal slip system in the α₂ phase are both very small and difficult to activate at room temperature [17,68,73], this study focuses on analyzing the potentially activatable slip systems in the B2 and α₂ phases under room-temperature conditions. Moreover, due to the difficulty in fabricating O-phase single crystals, the CRSS values for this phase have not been reported. In this study, the CRSS ratios of O-phase slip systems in Ti-22Al-25Nb alloys, as reported by Zhang et al. [60], are used as a reference. The ratios for basal slip, prismatic slip, first-order pyramidal slip, and second-order pyramidal slip are 1.19:1:1.37:1.39, respectively. The CRSS ratios indicate that the activation of prismatic and basal slip systems holds a significant advantage over the two types of pyramidal slip systems, with first-order pyramidal slip being more easily activated than second-order pyramidal slip.

A comparative analysis demonstrates that in the strongly textured B2-phase microstructure, the Schmid factors for both slip systems exceed 0.45 in different loading directions while showing nearly identical values. Consequently, the yield strength ratio between the RD and TD approaches 1. This indicates that the slip deformation behavior of the B2 phase exhibits negligible dependence on loading direction, contributing minimally to the yield strength anisotropy between the RD and TD, which is beneficial for plastic deformation.

For the α₂ phase, the prismatic slip system {10$\overline{1}$0}<11$\overline{2}$0> is preferentially activated in the RD direction due to its strong texture, while both basal and prismatic slip systems exhibit relatively low activation tendencies in the TD. According to Kishida et al. [74], the CRSS ratio between basal and prismatic slip systems reaches approximately 3.26, which would theoretically result in a yield strength ratio of about 0.15 between the RD and TD directions. However, experimental observations contradict this theoretical prediction due to the intrinsic brittleness of the α₂ phase. The limited plastic deformability stems from two intrinsic characteristics of the α₂ phase: (i) an insufficient number of available slip systems, and (ii) severely constrained dislocation mobility, as evidenced by extremely short slip lines confined within individual grain boundaries even when slip activation occurs [68]. These inherent limitations fundamentally restrict the plastic deformation capacity of the α₂ phase, rendering slip system activation largely ineffective in enhancing macroscopic ductility.

The O phase can be considered to have two orientations: the strongly textured O phase (referred to as O1) and the moderately textured O phase (referred to as O2). These two orientations are in a 90°/ND rotational relationship, as shown in Figure 10c. Due to this relationship, the activation of slip systems in the TD for O2 mirrors that in the RD for O1, and vice versa, as illustrated in Figure 13b,d. Consequently, the average Schmid factor of O2 in the TD can be directly represented by the average Schmid factor of O1 in the RD for the same slip system. Based on Figure 13, it can be observed that the most easily activated slip systems for the two orientations of the O phase in the RD are the prismatic {110}<1$\overline{1}$0> and the pyramidal {041}<114> slip systems, respectively. EBSD analysis revealed that the phase fractions of O1 and O2 are 86.8% and 13.2%, respectively. Assuming the CRSS for basal slip to be x and for prismatic slip to be 1.37x (the coefficient 1.37 is the ratio of CRSS of the two slip systems), the yield strengths in the rolling direction (${\mathsf{\sigma }}_{\mathrm{RD}}$) and transverse direction (${\mathsf{\sigma }}_{\mathrm{TD}}$) can be expressed as:

$${\mathsf{\sigma }}_{\mathrm{RD}}=86.8\%{\displaystyle \frac{\mathrm{x}}{0.46}}+13.2\% \times {\displaystyle \frac{1.37\mathrm{x}}{0.47}}=2.27\mathrm{x},$$

(3)

$${\mathsf{\sigma }}_{\mathrm{TD}}=86.8\% \times {\displaystyle \frac{1.37\mathrm{x}}{0.47}}+13.2\%{\displaystyle \frac{\mathrm{x}}{0.47}}=2.82\mathrm{x},$$

(4)

where 0.46 and 0.47 are the maximum Schmid factors of x and y, respectively. From Equations (3) and (4), it can be seen that the yield strength in the TD is higher than that in the RD. Based on the actual yield strengths (σ_RD = 936 MPa, σ_TD = 1136 MPa), the calculated values of x were 412.3 and 402.8, respectively. The close agreement between these values indicates the validity of the yield strength calculation method employed in this study. A higher yield strength corresponds to lower ductility; thus, the TD exhibits inferior ductility. In the RD, prismatic slip systems with a lower CRSS are more easily activated, whereas in the TD, pyramidal slip systems with a higher CRSS are required for activation. Consequently, the textured O phase contributes to greater elongation in the RD but higher yield strength in the TD. In summary, for Ti₂AlNb-based alloy sheets, the B2 phase exhibits superior deformability compared to both the α₂ and O phases. The crystallographic texture induces orientation-dependent activation behavior of slip systems in the α₂ and O phases, with the texture of the O phase being primarily responsible for the observed anisotropy in tensile properties.

5. Conclusions

This study analyzed the textures, the mechanisms underlying the formation mechanisms, as well as the microstructural evolution and phase transformations during heat treatment of rolled Ti₂AlNb-based alloy. Additionally, the influence of textures on tensile properties of the material was investigated. The key conclusions drawn from this study are as follows:

• In the rolled Ti₂AlNb-based alloy sheet, all three phases exhibit significant textures. The BCC phase develops a near-cubic texture; the α₂ phase is characterized by textures such as [0001]//TD, <$\overline{1}$2$\overline{1}$0>//ND, and {11$\overline{2}$0}<01$\overline{1}$0>; and the O phase exhibits a strong {100}<010> rolling texture. During heat treatment, the texture of the BCC phase weakens due to recovery and recrystallization, the texture of the α₂ phase remains largely unaffected, while the intensity of the O-phase texture shows a weakening trend with a decreasing phase content.
• The α₂ phase exhibits a tendency to spheroidize at 1000 °C, driven by the mechanisms involving grain boundary separation and recrystallization. During heat treatment, the O phase undergoes coarsening. When heat treatment is performed in or near the single-phase region, coarsening occurs due to the transformation from B2 to O phase. In contrast, when heat treatment is conducted in the two-phase region, coarsening is governed by the Ostwald ripening mechanism.
• The {11$\overline{2}$0}<01$\overline{1}$0> texture of the α₂ phase is a hot deformation texture formed during the rolling process. The strong texture of the O phase is influenced by both the texture of the B2 phase and the variant selection effect during its formation. Each B2 grain can only precipitate four variants, with O1 and O2 in the same B2-phase grain each comprising two variants.
• The Ti₂AlNb-based alloy sheet demonstrates enhanced ductility in the RD and superior yield strength in the TD, reaching a maximum value of 1136 MPa. The anisotropy in its tensile properties is primarily attributed to the difference in CRSS values between the prismatic slip system and the first-order pyramidal slip system of the O phase.