Effects of rolling and heat treatment on the microstructure and mechanical properties of newly designed high-temperature titanium alloy in plate manufacturing

Abstract

The alloy SJ1100, independently developed by Zhejiang Shenji Titanium Industry Co., Ltd, is a new type of lightweight, high-strength, high-elasticity titanium alloy with independent intellectual property rights. Rolled SJ1100 titanium-alloy plates and sheets were manufactured with different processes and parameters, and their microstructure and uniaxial tensile properties were characterized and compared by using metallography, scanning electron microscope, electron backscattering diffraction and tensile tests. The 2-mm sheet annealed at 820 °C for 40 min and straightened at 600 °C for 4 h exhibited the highest yield strength and ultimate tensile strength as well as the lowest ductility, while the 30-mm one annealed at 820 °C for 40 min and aged at 600 °C for 16 h exhibited the lowest yield strength and ultimate tensile strength with the highest ductility. Straightening is thought to be more deleterious to ductility compared with aging. Annealing after straightening decreases the average grain size, lowers the intensity level of texture, and relieves the residual stress of the sheet. A larger plastic deformation results in an energetically more unstable regime, which makes the sheet easier to recrystallize during annealing.

1 Introduction

TA11 is a near-α high-temperature titanium alloy, primarily composed of Ti-8Al-1Mo-1V (in weight). This material possesses high strength, excellent thermal stability, and high-temperature creep resistance (Sun, 2016). This type of material is primarily used in high-speed rotating components such as aircraft engine compressor discs and rotor blades. Since aircraft components are subjected to harsh working environments over extended periods, they are required to possess excellent mechanical properties at both room and high temperatures, with particularly stringent demands on the thermal stability and high-temperature creep performance, so as to ensure the service life of the engine (Zheng et al., 2017; Duan et al., 2016). As a β-stabilizer, Mo suppresses the transformation of the β phase, refines the microstructure, reduces grain size, enhances creep resistance and thermal stability, and works synergistically with V to balance the alloy’s strength and plasticity. This renders TA11 good weldability and thermal workability (Zhang, 2024; Lu et al., 2012). When the content is 1.5%, the strength and hardness reach their peak values (Zhang, 2024).

It is noticed that TA11 has some disadvantages, such as the limited strength (normally <1000 MPa) and operating temperature (normally ∼450 °C), a slightly high density (∼4.47 g/cm³), and a limited hot workability (Zheng et al., 2017; Duan et al., 2016; Wang and Zheng, 2025; Ying et al., 2025). The alloy SJ1100, independently developed by Zhejiang Shenji Titanium Industry Co., Ltd, is a new type of lightweight, high-strength, high-elasticity titanium alloy with independent intellectual property rights (Liu, 2024). The nominal composition is Ti-8Al-2V-1Cr-1Zr (wt%). SJ1100 is an α-β dual-phase alloy with excellent overall properties, stable microstructure, superior toughness, plasticity, and high-temperature deformation capabilities. It responds well to hot pressing and can be strengthened through quenching and aging. The alloy is easy to process, and readily producible, making it suitable for various application fields requiring low-density, high-strength titanium alloys. Al is selected as the main element to improve the strength of this titanium alloy, stabilizing α phase and enhancing high-temperature strength while reducing the density. In order to balance the reduced workability caused by the increase in Al content, a small amount of V and Cr is added to the alloy. V has the same crystal structure as Ti, and is able to inhibit grain nucleation and refine grains, and to ensure the alloy’s high plasticity and high strength. At the same time, the omission of noble metals like Mo lowers the cost of the alloy. A small amount of Cr and Zr is added to improve its strength and oxidation resistance. Particularly, Cr mainly plays a role in solid solution strengthening, enhancing the alloy’s plasticity, toughness, workability, hardenability, quenching property, and creep resistance, and can also refine the cast microstructure, but deteriorates the thermal stability performance (Xin et al., 2008; Wu et al., 2017). When the concentration reaches up to 6%, no α′ phase is formed after solid solution treatment and quenching, forming fine α” phase instead, and hardness reaches a peak. After aging, α” phase becomes nucleation sites for α phase, with α grains reduced to sub-micron level, and hardness reaches a peak (Wu et al., 2017). Zr is a neutral element that can reduce grain size by inhibiting the growth of α phase due to its slow diffusivity (Xie et al., 2024; Jiang et al., 2019), and strengthen titanium alloys as substitutional atoms through solid solution, while precisely regulating phase evolution: suppressing α” phase formation during quenching, stabilizing α/β phases, increasing residual β-phase content, reducing martensitic transformation temperature, and inhibiting ω-phase precipitation during aging. Additionally, Zr boosts oxidation resistance by inducing a dense columnar oxide film, optimizes mechanical properties by lowering elastic modulus and hardness while improving ductility and tensile strength, and enhances workability with reduced anisotropy. These properties make Zr a critical element for high-performance titanium alloy design (Fu et al., 2015; Fei and Wang, 2024; Lian et al., 2013; Xie et al., 2010). The synergistic effect of Cr and Zr maintains high strength, low density, high plasticity, and excellent processing performance of the alloy. Ultimately, the alloy’s density is reduced to 4.35 g/cm³. The tensile strength exceeds 1150 MPa, and the elongation reaches over 10%. This achieves a good trade-off of strength and plasticity which is rarely found in existing technical materials, with the characteristics of being lightweight and easy to produce and process. It is particularly necessary to study the effects of hot working and heat treatment on the microstructure and mechanical properties of SJ1100 titanium alloy plates and sheets.

In order to understand the influence of different processes and parameters on the microstructure and mechanical properties, it would be instructive to examine SJ1100 alloy sheets/plates produced by different processes and parameters after rolling. This study focuses on the microstructure, texture, and tensile properties of three sheets and a plate processed under the combination of straightening, aging, and annealing after rolling.

2 Materials and methods

Sheets/plates of SJ1100 alloy with a nominal composition of Ti-8Al-2V-1Cr-1Zr (wt%) were produced at Zhejiang Shenji Titanium Industry Co., Ltd. In comparison with TA11, V content is increased to 2 wt%, and 1 wt% of Mo is replaced by 1 wt% of Cr and 1 wt% of Zr. Raw materials such as titanium sponge, aluminum beans, and aluminum-vanadium master alloy were blended and first pressed to produce electrodes. A cylindrical ingot was then produced by vacuum arc remelting for three times. The β-transus temperature was measured to be 967 °C. The ingot was forged above β-transus temperature and cut into rectangular slabs. The usable portion of the resulting slabs was hot-rolled till a thickness of 30 mm. One of the four different samples (#4) was obtained from one of the 30-mm-thick plates. The rest of the plates were further processed with commuting rolling and cold-rolling till a thickness of 2 mm. Three samples were obtained from the as-rolled 2-mm-thick sheets (#1∼3), which were first annealed at 820 °C for 40 min. Then sample #1 was obtained from the sheets straightened under a load of 5 tons at 600 °C for 4 h. In comparison, sample #2 was obtained from the sheets aged at 600 °C for 16 h. To illustrate the influence of annealing, sample #3 was annealed at 800 °C for 50 min on the basis of sample #1. Sample #4 was annealed at 820 °C for 40 min and aged at 600 °C for 16 h. The annealing temperatures of 800 °C and 820 °C lie within the α+β two-phase field, which favor equiaxed α formation. Table 1 summarizes the heat-treatment processes and parameters of the samples examined in this study.

One slice of specimen for optical microscopy was electro-discharge machined (EDM) from each of the samples processed with different heat treatment processes, and was mounted in thermoset, ground using standard 400, 800, 1500, 2000, and 2500 grit abrasive discs and polished using 2.5, 1.0, and 0.5 μm alumina solution. The polished surface was etched using a solution of 12 mL of nitric acid and 5 mL of hydrofluoric acid dissolved in 83 mL of distilled water for 5∼8 s to reveal different phases. The microstructure was examined using an optical microscope and a Thermo Fisher Apreo 2s field emission scanning electron microscope (FE-SEM). Electron backscattering diffraction (EBSD) specimens EDM’ed from the four samples were polished both mechanically and electrolytically. After mechanical polishing as described above, the surface was electrolytically polished using a solution of 20 mL of perchloric acid dissolved in 380 mL of methanol at ∼ -30 °C and using 30V and 0.3A for 30 s. The Oxford C-Swift + EBSD probe and AZtecCrystal software were used to acquire specimen data information and analyze the specimen grain size, grain boundary distribution, dislocation density, and micro-texture.

Uniaxial tensile tests for 2-mm sheets were performed using flat, dog-bone shaped specimens with a gauge length of ∼65 mm and a gauge width of 12.5 mm, and those for 30-mm plate were performed using hourglass-shaped specimens with a gauge length of ∼65 mm and a gauge diameter of 10 mm. These were electro-discharge machined along both the rolling direction and the transverse direction. All the tensile tests were performed at room temperature using an MTS electronic universal testing machine and were in accordance with ASTM B265. The maximum load of the tensile machine is 300 kN, with the strain rate controlled at 0.005/min for the first segment and 0.4/min for the second one; the extensometer gauge lengths for the dog-bone shaped and hourglass-shaped specimens are 30 and 50 mm, respectively, with a range of 10 mm and an accuracy level of 0.5.

3 Results

3.1 Metallography and microstructure

Representative optical micrographs for the four heat-treated specimens are shown in Figure 1. In all four groups of specimens, the intergranular β phase between the α grains can be observed (the darker structure among the white α phase in Figure 1). This is due to the addition of 1 wt% of V and 1 wt% of Zr as β stabilizing elements in SJ1100 titanium alloy, which retains a small amount of β phase at room temperature. The alloy consists of equiaxed and lamellar α phase and β phase. This is due to the SJ1100 titanium alloy undergoing recovery and incomplete recrystallization process during annealing at 820 °C for 40 min, primarily occurring through polygonization to form subgrain boundaries, with their microstructure inheriting the rolled state morphology, such as the equiaxed α phase formed by dynamic recrystallization and the deformed flattened lamellar α phase. α and β grains are aligned parallel to the transverse direction in each specimen. The grain distribution of Samples #1 and #2 is more uniform, exhibiting a sufficiently recrystallized equiaxed grain microstructure, as a result of sufficient plastic deformation prior to heat treatment; the grain boundaries of Sample #3 are more pronounced due to the additional annealing; the grains of Sample #4 exhibit an elongated strip-like shape, with evident features of deformation flow lines, as a result of insufficient plastic deformation prior to aging and the absence of commuting rolling.

FIGURE 1

Figure 2 shows the secondary electron imaging of scanning electron microscope for four sets of specimens. Different from the colors observed in optical microscope, the bright structure among the dark α phase represents β phase. Although most of the grain boundaries of matrix α phase are not well exposed, it can be preliminarily seen by the distribution of β phase that some α grains are slightly stretched along the transverse direction in each specimen, indicating the inheritance of the microstructure from the rolled state. As below, EBSD is employed to further characterize the features of the specimens including grain size and texture. In Sample #1 the bright β phase is the smallest in size and most uniformly distributed, which generally corresponds to higher strength. In Sample #2, due to the effect of elongated time at 600 °C (16 h vs. 4 h), the β phase begins to enlarge, with a tendency to agglomerate, compared with those in Sample #1. The changes in Sample #3 are even more apparent, as the β phase has formed small blocks rather than particles after the additional annealing, and the distribution is less uniform. Sample #4 exhibits the most significant microstructural difference, with the β phase forming a continuous network that segments the matrix, as a result of insufficient plastic deformation and the absence of commuting rolling, preserving inadequately broken β phase.

FIGURE 2

EBSD was conducted at the center of transverse section of each specimen, with a scanned area of about 50 μm × 34 μm and a step size of 0.05 μm. The Oxford AZtec Crystal software was used to analyze the α grain orientation distribution map as shown in Figure 3. Since the proportion of intergranular β phase is very low and the size is small, its contribution to the orientation distribution was neglected during the EBSD scanning. This experimental strategy does not affect the analysis of actual matrix grain size, grain boundaries, and lattice distortion within grains of α phase. The elongated plate-like structure as seen in Figure 1 has actually fragmented into a large number of small grains, but the small grains are aligned along the transverse direction. Since there is no phase (shown as black in the metallography) between the small grains, during sample preparation, the etchant can only expose the phase boundaries but not the grain boundaries. Therefore, in the metallography, it appears as an elongated plate-like structure. As shown in Figures 3a–d, the microstructures of the four specimens are equiaxed α grains, with β phase distributed in grain boundaries, slightly stretched along transverse directions after annealing, with diverse grain orientations (different colors represent grains with different orientations). Sample #1 exhibits a high proportion of {0001} orientation (red grains), indicating that annealing at 820 °C and straightening at 600 °C enables α grains to rapidly grow in the basal {0001} orientation during the process. This is significantly different from the cylindrical texture observed in the other 3 samples ({1010}//rolling plane//close-packed plane, 〈1120〉//rolling direction//close-packed direction) formed during the post-rolling heat-treatment process.

FIGURE 3

For the EBSD data above, high-angle grain boundaries were defined as >10°, and regions with more than 100 pixels (corresponding to a grain area of 0.25 µm²) were identified as grains. Since the volume fraction of β phase is negligible and the effect of α grains on the properties of the plates is dominant, the statistical results of α grains are discussed thereafter. In each heat-treated specimen, 1300∼1700 grains were counted, which is adequate to form statistical patterns (fine subgrains are included in the statistics). The average diameter of each grain was determined by using the equivalent circle diameter method to calculate the average grain size. As shown in Figure 4, the average grain diameter was the largest of 2.9 µm after annealing at 820 °C for 40 min and straightening at 600 °C for 4 h (Specimen #1), while that was the smallest after sample #4 was annealed at 820 °C for 40 min and aged at 600 °C for 16 h (2.51 µm). Such a difference reflects the effect of the driving force of recrystallization and grain growth. Straightening introduces local plastic strain, which induces recrystallization nucleation, resulting in the largest average grain size in Sample #1 among the four. On the contrary, with the lowest degree of plastic deformation, Sample #4 stored the least strain energy and exhibited the smallest average grain size.

FIGURE 4

3.2 Crystallographic texture

Figure 5 presents the pole figures of α grains in SJ1100 samples under various heat treatment conditions, derived from EBSD data, along the 〈0001〉, <1 10>, and <01 0> crystallographic directions. When annealed at 820 °C for 40 min followed by straightening at 600 °C for 4 h, the {01 0} plane of the strongest texture is parallel to the rolling plane and the <11 3> orientation is parallel to the rolling direction, with a maximum texture intensity of 9.81 (Figure 5a). Cold rolling with a high deformation degree activates the {10 1}<11 3> and {11 2}<11 3> slip systems, forming a deformation texture centered on the {01 0}<11 3> texture (Zigan et al., 2020). When annealed at 820 °C for 40 min followed by aging at 600 °C for 16 h, the {02 1} plane of the strongest texture is parallel to the rolling plane and the < 012> orientation is parallel to the rolling direction, with a maximum texture intensity of 11.09 (Figure 5b). Cold rolling forms the initial deformation texture; during aging, the {02 1}< 012> texture orientation with low grain boundary energy preferentially grows. Diffusion-dominated selective growth strengthens the {02 1}< 012> texture (Wang and Huang, 2003). When annealed at 820 °C for 40 min followed by straightening at 600 °C for 4 h and annealing at 800 °C for 50 min, the {25 3} plane of the strongest texture is parallel to the rolling plane and the < 26> orientation is parallel to the rolling direction, with a maximum texture intensity of 9.61 (Figure 5c). {25 3}< 26> texture is similar to the {11 2}<11 3> texture, a composite texture after high-temperature straightening and secondary annealing, with grain orientations showing complex multi-directional preference (Zigan et al., 2020). When sample #4 was annealed at 820 °C for 40 min followed by aging at 600 °C for 16 h, the { 2 1} plane of the strongest texture is parallel to the rolling plane and the < 8 12> orientation is parallel to the rolling direction, with a maximum texture intensity of 7.23 (Figure 5d). During annealing at 820 °C, the grains of the titanium alloy undergo recrystallization, and the grain orientation is influenced by the initial hot-rolling texture. Meanwhile, during aging at 600 °C, the nucleation and growth of precipitates have a weak adjusting effect on the grain orientation. Ultimately, a texture similar to { 2 1} <01 3> is formed, which is a typical weak texture after hot-rolling, annealing, and aging.

FIGURE 5

3.3 Disorientation angles

The distributions of random pairs (i.e., 10,000 random orientation pairs) and neighboring pairs (e.g., grain boundaries in the microstructure) of disorientation angles were plotted for the four samples as statistical histograms in Figure 6. The consistency percentage of the random pairs with the theoretical distribution was calculated automatically by the software as the integral goodness-of-fit between the actual random-pair distribution curve and the theoretically completely random distribution curve, which can be used as a criterion for texture strength. The peaks in the disorientation angle distribution histogram of neighboring pairs correspond to certain specific rotation angles, such as those between twins and the matrix, and are related to adjacent grains. The disorientation angle distributions of random pairs show low consistency with those of neighboring pairs, indicating that the disorientation distribution of neighboring pairs is significantly affected by subgrain boundaries. The strongest peak of each sample appears at the low-angle end (<10°), which is related to subgrain boundaries. Figure 6 also shows peaks corresponding to adjacent pairs at disorientation angles in the range of 60°∼65°, which is indicative of a significant proportion of {1122}〈1123〉 compressive twin boundaries (Chun et al., 2005; Wang et al., 2018; Liu et al., 2018; Beladi et al., 2014; Yu et al., 2024). The congruence between random pairs and theoretical values of Sample #1 (92%) is apparently lower than those of the other 3 samples (97∼98%), which indicates a preferred orientation and a higher texture intensity in Sample #1. Such texture intensity is lowered by prolonged aging time (16 h compared with 4 h) and additional annealing in Samples #2 and #3, respectively. In Sample #4, the texture intensity is not developed due to the insufficient deformation degree.

FIGURE 6

3.4 Kernel average misorientation

Figure 7 shows the kernel average misorientation (KAM) distribution maps of the aforementioned sample groups, with a calculation range of 0°–5°, where blue corresponds to the minimum value of 0° and red corresponds to the maximum value of 5°. Kernel Average Misorientation is a method in EBSD data analysis that characterizes local misorientation angles based on central pixels, and this indicates the distribution of local strain at the grain boundaries of crystalline materials. The KAM value of each data point is the average misorientation of all other data points within a certain radius around the central data point, which can reflect the density of geometrically necessary dislocations (GNDs) in the crystalline material. For Samples #2 and #3 (Figures 7b,c), most grain interiors appear blue, indicating low local strain and low dislocation density within the grains, which is attributed to the prolonged aging time and additional annealing, respectively. In comparison, in Samples #1 and #4 (Figures 7a,d), more grains exhibit higher local strain.

FIGURE 7

3.5 Tensile behavior

Table 2 and Figure 8 show the room temperature tensile properties of the SJ1100 samples heat treated with different processes. When annealed at 820 °C for 40 min followed by straightening at 600 °C for 4 h, the sample exhibited the highest tensile strength (1154∼1221 MPa) and yield strength (1137∼1153 MPa), and the lowest elongation (1.5∼6.5%), which is attributed to the higher deformation degree and the shortest heat treatment time. In contrast, after sample #4 was annealed at 820 °C for 40 min followed by aging at 600 °C for 16 h, it showed the lowest yield strength (979∼1022 MPa) and tensile strength (1010∼1083 MPa) due to the insufficient plastic deformation. Overall, the SJ1100 sample annealed at 820 °C for 40 min followed by straightening at 600 °C for 4 h and annealing at 800 °C for 50 min had a good balance of strength (1076∼1106 MPa) and ductility (12.5∼13%), demonstrating excellent overall performance.

FIGURE 8

4 Discussion

4.1 Comparison between straightening and aging

It can be seen in Figure 4 that the sheet annealed at 820 °C for 40 min and straightened at 600 °C for 4 h has a larger average grain diameter (2.9 µm) compared with that annealed at 820 °C for 40 min and aged at 600 °C for 16 h (2.69 µm). The main reason is that straightening introduces local plastic strain (as seen in Figure 7a), which induces recrystallization nucleation and results in non-uniform grain coarsening; whereas aging treatment pins the grain boundaries with precipitates, effectively inhibiting grain growth. Although straightening is a low-deformation process, the local stress can still cause non-uniform plastic strain in titanium alloy sheets, accumulating a large number of dislocations and stored energy. When the temperature reaches the critical point for recrystallization, these high-strain regions become heterogeneous nucleation sites, triggering localized recrystallization. Li et al. stated in (Chao et al., 2024) that some meta-dynamic recrystallization grains precipitate near the grain boundaries of original β phases in TC18, resulting in increasing of dynamic recrystallization fraction. Due to non-uniform deformation distribution in SJ1100, the recrystallized grains coarsen and grow in strain-concentrated areas, forming a microstructure with non-uniform sizes and a larger statistical average grain size. α-β titanium alloys are prone to develop ‘necklace-type’ recrystallized grains during straightening, with actual matrix grain sizes significantly larger than the original grains. Aging treatment precipitates secondary phases in the matrix phases. These precipitates are uniformly dispersed along grain boundaries and within grains, generating a strong Zener pinning effect and significantly increasing the energy barrier for grain boundary migration. This mechanism effectively prevents grain coarsening and maintains a fine, uniform microstructure.

It can be seen from Table 2 that the specimen annealed at 820 °C for 40 min and straightened at 600 °C for 4 h exhibited a lower ductility (1.5∼6.5%) compared with that annealed at 820 °C for 40 min and aged at 600 °C for 16 h (5∼8.5%). Straightening is carried out at a high temperature. Influenced by external forces, creep occurs during the straightening process, and a large number of dislocations in titanium alloys (especially the α phase) slip and reorganize, forming textures or preferred orientations. The texture makes the plastic deformation capability highly anisotropic, as shown by the anisotropy ratio farthest away from 1 among the four specimens listed in the last column of Table 2. At the same time, dislocations pile up at grain boundaries and phase boundaries, forming a high-density substructure. Although this partially reduces residual stress, it does not achieve structural homogenization, causing early stress concentration during tension and the preferential initiation of microcracks at texture interfaces. As shown in Figure 7a, the kernel average misorientation (KAM) obtained from the EBSD data reflects the localized strain distribution. The strain is concentrated in the grain boundaries. This significant strain imbalance results in low ductility. In addition, a larger statistical average grain size also reduces the plasticity. The aging process does not have a significant impact on the fraction and morphology of the primary α phase in the microstructure. In this process, the primary α phase has an equiaxed morphology, and the equiaxed α phase exhibits good compatibility during deformation, facilitating the initiation of slip and resulting in better plasticity of the alloy (Yang et al., 2024). Straightening is a process forced mechanically. The essence of ‘straightening’ lies in eliminating macroscopic bending through local creep, but at the cost of texture locking in the microstructure. In contrast, aging is a thermodynamically driven organizational optimization process. It achieves a plasticity through the strain coordination effect of precipitates. Experiments show that after aging at 540 °C for 4 h, the elongation of TC4 titanium alloy can reach 13.4% (Fan and Feng, 2018), whereas the elongation of the hot-straightened state at the same temperature generally remains below 10% (Yao et al., 2018). Fracture morphology analysis for SJ1100 will be carried out in future work.

4.2 Additional annealing

Additional annealing after straightening (Sample #3) reduced the average grain size and texture intensity, as seen in Figures 4, 5, respectively. The anomalous phenomenon of the reduction of statistical average grain size is due to the effect of high-density recrystallization nucleation driven by deformation substructures. Although straightening is carried out at a high temperature, it leads to the formation of high-density dislocation entanglements, cellular substructures, and low-angle grain boundaries in the α phase. These structures serve as preferred nucleation sites during the initial stage of annealing, significantly increasing the nucleation rate. According to recrystallization theory, nucleation density is positively correlated with stored energy. The high stored energy state of the thermally straightened condition produces a large number of fine, equiaxed nuclei during annealing, far exceeding the rate of normal grain growth, thereby achieving a ‘nucleation-dominated refinement’ before the grains have time to grow. Yu et al. reported in (Feilong et al., 2022) that during short-period annealing of Ti–6Al–4V sheets, the pinning effect of fine β phase particles and the competition mechanism between α and β grains inhibited the coarsening of grains; meanwhile, partial recrystallization and recovery make the combination of micro-grains and ultrafine-grains. Straightening under stress of SJ1100 induces the formation of a strong texture, causing some grains to have highly consistent orientations, resulting in low-energy grain boundaries (low-angle grain boundaries or special Σ grain boundaries). The migration rate of these grain boundaries is significantly lower than that of high-angle grain boundaries, leading to slow or even stagnant growth during annealing. Meanwhile, grains with other orientations, although nucleating actively, have their overall grain size limited due to texture competition, forming a dual refinement mechanism of ‘preferential nucleation and preferential suppression’. It can be seen in Figures 6a,c that the low-angle grain boundary peak decreased and the fraction of high-angle grain boundaries (60∼65°) increased after additional annealing, indicating that subgrain coalescence and recrystallization nucleation also play an important role in grain refinement.

The texture strength of titanium alloy sheet subjected to straightening decreases after annealing, which is the result of the combined effect of a dual mechanism: recrystallization nucleation dominance and orientation randomization. Li et al. investigated the effect of annealing treatment on the microstructural evolution of titanium and concluded that as the annealing temperature increases, the texture strength gradually decreases (Li et al., 2025). Straightening generates a large number of dislocation entanglements and subgrain boundaries at high temperatures, forming a high-energy storage state. During annealing, these substructures become preferred nucleation sites, leading to a sharp increase in recrystallization nucleation density. The orientations of the newly formed grains are randomly distributed, quickly consuming the originally strongly oriented grains, causing the overall crystal orientation to transition from highly ordered to statistically disordered, and reducing texture intensity. Experiments show that after annealing at 800 °C, the texture strength of TC4 titanium alloy decreases to its lowest, dropping by about 40%–60% compared to the as-rolled state (Wang et al., 2023). In the annealed SJ1100 the orientation distribution of recrystallized grains tends to be random, the peak intensity in the pole figure (Figure 5c) is weakened, the orientation concentration is reduced, achieving texture randomization. The weakening of texture after annealing is not a simple ‘elimination’, but an inevitable result of microstructural reconstruction: the crystallographic order established by straightening is disrupted by the ‘reset’ mechanism of recrystallization under thermal activation, while the precipitated phases ensure that this weakened state is irreversible. From the special grain boundary distribution map in Figures 9a,b, we can see the difference in the distribution of different types of grain boundaries before and after additional annealing. After additional annealing, the <7 10 0 > 90° grain boundaries (yellow lines in Figures 9a–c) increased from 0.47% to 1.24%. Combined with the rotated axis distribution (Figure 9d), the ideal Burgers relationship in the recrystallization can be expressed as <7 10 0 > 90°. This shows that the formation of corresponding variants can effectively relieve the residue stress in the straightened sheet. This process is a key heat treatment path for achieving isotropic plasticity and optimized formability in titanium alloys, and is widely used in the final heat treatment processes of aerospace structural components.

FIGURE 9

After annealing, the residual stress inside straightened titanium alloy sheet is significantly reduced (as seen in Figure 7c), resulting in higher ductility, and the core mechanism stems from thermally activated microstructural rearrangement and stress relaxation processes. A numerical simulation study was conducted on the stress-relief annealing heat treatment of a titanium alloy frame component, and concluded that the component experienced the release and redistribution of internal stress during the heat treatment process (Jihao et al., 2024). Straightening introduces high-density dislocation entanglements and subgrain structures at high temperatures, creating a non-uniform plastic strain field and resulting in residual stresses of up to several hundred MPa. During annealing, atomic thermal motion is enhanced, and dislocations achieve directional movement through slip, climb, and cross-slip, with neighboring dislocations undergoing annihilation, leading to a significant reduction in dislocation density. This process directly eliminates lattice distortion energy and is the primary driving force for the release of residual stress. In the straightened state, a subgrain structure composed of a large number of low-angle grain boundaries (<15°) gradually merges into high-angle grain boundaries during annealing through grain boundary migration and orientation adjustment. This process is accompanied by local reorientation of the grains, which helps ‘flatten’ originally concentrated local stress fields, thereby eliminating macroscopic stress gradients. Grain boundaries, as sources of stress concentration, release elastic strain energy during the relaxation process, further reducing the overall internal stress level. The precipitates precipitate during annealing, and their main role is to inhibit abnormally large grain growth after recrystallization, thereby preventing stress from reaccumulating. This mechanism does not participate in the release of stress itself, but by maintaining fine grains and a uniform microstructure, it allows the material to maintain a long-term low residual stress state after annealing, providing a stable effect of ‘locking’ stress reduction. The reduction of residual stress after annealing is not the result of a single mechanism, but rather the combined effect of thermally activated plastic relaxation and microstructural evolution. This process is widely used in the aerospace industry to improve the dimensional stability and fatigue life of titanium alloy components and is a key step in the heat treatment process of critical structural parts.

4.3 Plate thickness

The sheet with greater deformation (Sample #2) resulted in an average grain size larger than that of the plate (Sample #4) as well as a lower local strain (comparing Figures 7b,d). The essence of this phenomenon is that highly deformed sheet achieved complete recrystallization (Li et al., 2025), while low-deformed plate exhibited insufficient recrystallization due to non-uniform heat conduction and pinning by precipitates, retaining the original grains, resulting in an anomalous result in terms of average grain size (Feilong et al., 2022). It can be systematically explained from two aspects: the driving force of recrystallization and geometric effects. Plastic deformation introduces a high density of dislocations, significantly increasing the stored energy inside the material (the driving force for recrystallization). When the degree of deformation is large, the nucleation rate rises sharply, forming a large number of recrystallization nuclei, which lays the foundation for rapid grain growth. When the deformation is small, the dislocation distribution is non-uniform, the nucleation rate is low, and recrystallization occurs only in localized areas. Grain growth lacks sufficient driving force. A high surface area-to-volume ratio in sheets allows for uniform heat transfer, resulting in small temperature gradients during heat treatment. The entire material simultaneously reaches the recrystallization temperature, achieving uniform and complete recrystallization, with grains able to grow freely. In comparison, in plates, heat conduction is slow, and the temperature in the center lags behind the surface. When the soak time is insufficient, only partial recrystallization occurs in the surface layer, while the center still retains deformed microstructure, resulting in an overall suppression of the statistical average grain size.

5 Conclusion

References

• 1

  BeladiH.ChaoQ.RohrerS. G. (2014). Variant selection and intervariant crystallographic planes distribution in martensite in a Ti–6Al–4V alloy. Acta Mater.80, 80478–80489. 10.1016/j.actamat.2014.06.064

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  ChaoLiZhangS.LinY. (2024). Microstructure and texture evolution law of TC18 titanium alloy during double-pass hot deformation process. Forg. and Stamp. Technol.49 (07), 81–89. 10.13330/j.issn.1000-3940.2024.07.009

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  ChunY.YuS.SemiatinS.HwangS. (2005). Effect of deformation twinning on microstructure and texture evolution during cold rolling of CP-titanium. Mater. Sci. and Eng. A398 (1-2), 209–219. 10.1016/j.msea.2005.03.019

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  DuanX.LiW.LiF.WangS. (2016). Effects of hot working processes on the microstructure and properties of TA11 titanium alloy bar. World Nonferrous Met. (14), 47–49.

  • Google Scholar
  • View reference in article

• 5

  FanZ.FengH. (2018). Study on selective laser melting and heat treatment of Ti-6Al-4V alloy. Results Phys.10, 10660–10664. 10.1016/j.rinp.2018.07.008

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  FeiJ.WangX. (2024). Effect of Zr and O solution strengthening on the anisotropy of mechanical properties of Ti alloys. J. Mater. Sci. and Eng.42 (06), 888–896+904. 10.14136/j.cnki.issn1673-2812.2024.06.002

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  FeilongY.YunZ.CharlieK.HailiangY. (2022). High strength and toughness of Ti–6Al–4V sheets via cryorolling and short-period annealing. Mater. Sci. and Eng.A, 854. 10.1016/J.MSEA.2022.143766

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  FuB.WangH.ZouC.WeiZ. (2015). The influence of Zr content on microstructure and precipitation of silicide in as-cast near α titanium alloys. Mater. Charact.99, 9917–9924. 10.1016/j.matchar.2014.09.015

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  JiangB.WenD.WangQ.CheJ.DongC.LiawP. K.et al (2019). Design of near-α Ti alloys via a cluster formula approach and their high-temperature oxidation resistance. J. Mater. Sci. and Technol.35 (6), 1008–1016. 10.1016/j.jmst.2018.12.013

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  JihaoY. I.MingyongC. A. I.LisenL. I.TangW.YeX.HuaC.et al (2024). Study on deformation of titanium alloy part by stress relief annealing. Hot Work. Technol.53 (04), 151–154. 10.14158/j.cnki.1001-3814.20221776

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  LiX.GuoY.ZhangS.WangL.FanX.TianY. (2025). Effect of cryogenic rolling and annealing treatment on microstructure evolution and mechanical properties of pure titanium plate. Titan. Ind. Prog.42 (02), 23–27. 10.13567/j.cnki.issn1009-9964.2025.02.001

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  LianC.YangS.ZhouH.TangR.LiD. (2013). Effect of alloying element (Nb,Ta,Fe and Zr) on microstructure and mechanical properties of biomedical titanium alloy. Hot Work. Technol.42 (14), 40–42. 10.14158/j.cnki.1001-3814.2013.14.032

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  LiuT. (2024). A Lightweight high-temperature Titanium Alloy and its Preparation Method. ChinaCN116426791B.

  • Google Scholar
  • View reference in article

• 14

  LiuN.WangY.HeW.LiJ.ChapuisA.LuanB. f.et al (2018). Microstructure and textural evolution during cold rolling and annealing of commercially pure titanium sheet. Trans. Nonferrous Metals Soc. China28 (6), 1123–1131. 10.1016/S1003-6326(18)64748-X

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  LuJ.ZhaoY.GeP.ZhouW. (2012). Research development of the effect of Mo element on precipitation behavior in titanium alloy. Titan. Ind. Prog.29 (02), 10–14. 10.13567/j.cnki.issn1009-9964.2012.02.010

  • CrossRef
  • Google Scholar

• 16

  SunQ. (2016). Effect of hot working process on properties of TA11 titanium alloy bars. Chem. Eng. Des. Commun.42 (12), 74+90.

  • Google Scholar

• 17

  WangY. N.HuangJ. C. (2003). Texture analysis in hexagonal materials. Mater. Chem. Phys.81 (01), 11–26. 10.1016/s0254-0584(03)00168-8

  • CrossRef
  • Google Scholar

• 18

  WangJ.ZhengB. (2025). Effect of strain rate on hot deformation behavior of TA11 titanium alloys. Titan. Ind. Prog.42 (03), 20–26. 10.13567/j.cnki.issn1009-9964.2025.03.009

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  WangY.HeW.LiuN.ChapuisA.LuanB.LiuQ. (2018). Effect of pre-annealing deformation on the recrystallized texture and grain boundary misorientation in commercial pure titanium. Mater. Charact.136, 1361–11. 10.1016/j.matchar.2017.11.059

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  WangW.ZhouS.GongP.ZhangH.ShiY.WangK. (2023). Effect of anneal treatment on microstructure, texture and mechanical properties of TC4 alloy plates. Chin. J. Mater. Res.37 (01), 70–80. 10.11901/1005.3093.2022.103

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  WuD.LiuL. B.ZhangL. G.ZengL. J.ShiX. (2017). Investigation of the influence of Cr on the microstructure and properties of Ti6Al4VxCr alloys with a combinatorial approach. J. Materi Eng Perform26, 4364–4372. 10.1007/s11665-017-2822-4

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  XieH.LouY.LiuS.ZhaoJ.SuG.XueS. (2010). Effects of Ta and Zr on properties and microstructure of solid solution Ti alloy. Special Cast. Nonferrous Alloys30 (02), 111–113+92. 10.3870/tzzz.2010.02.004

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  XieM.HuangS.WangZ.RamamurtyU. (2024). High-temperature fracture behavior of an α/β Titanium alloy manufactured using laser powder bed fusion. Acta Mater.277, 277120211–120211. 10.1016/J.ACTAMAT.2024.120211

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  XinS.ZhaoY.ZengW. (2008). Mechanism of V and Cr on mechanical properties of Ti40 burn resistant titanium alloy. Chin. J. Nonferrous Metals (07), 1216–1222. 10.19476/j.ysxb.1004.0609.2008.07.008

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  YangM.ZhangM.YuC.YueX.TongX.ZhangY. (2024). Effect of solution and aging treatment on microstructure and mechanical properties of TA10 titanium alloy. J. Metal. World (01), 49–52. 10.3969/j.issn.1000-6826.2022.05.1101

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  YaoX.FuL.ShanA. (2018). Microstructure and mechanical properties of ultrafine-grained TC4 titanium alloy prepared by severe hot rolling. Mater. Mech. Eng.42 (03), 57–61+66.

  • Google Scholar
  • View reference in article

• 27

  YingJ.XuS.LiuG.PanS.FanQ.ChengX. (2025). Enhanced dynamic compression properties of a low density near-α titanium alloy associated with deformation induced laminated microstructure and dynamic segregation. Acta Mater.285, 285120653-120653. 10.1016/J.ACTAMAT.2024.120653

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  YuX.YuanQ.ZhouQ.LiangB.ChenY.LiuT. (2024). Influence of annealing temperature on microstructure and mechanical properties of Gr.4 pure titanium sheet. Titan. Ind. Prog.41 (04), 17–22. 10.13567/j.cnki.issn1009-9964.20240826.001

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  ZhangJ. (2024). The influence of different mass fractions of Mo elements on powder metallurgy properties. Shanxi Metall.47 (10), 49–51. 10.16525/j.cnki.cn14-1167/tf.2024.10.017

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  ZhengN.ZhangY.HeY.ShangJ.HouF.LaiY. (2017). Influence of microstructure on mechanical properties of titanium alloy TA11 bar. Forg. and Stamp. Technol.42 (08), 146–151. 10.13330/j.issn.1000-3940.2017.08.028

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  ZiganW.PingY.XinfuG.OnukiY.SatoS. (2020). Transformation textures in pure titanium: texture memory vs surface effect. Mater. Charact.164, 110359. 10.1016/j.matchar.2020.110359

  • CrossRef
  • Google Scholar
  • View reference in article