Research Progress on the Creep Resistance of High-Temperature Titanium Alloys: A Review

Abstract

High-temperature titanium alloys are one of the most important research directions in the field of high-temperature aerospace alloys. They are mainly used in high-temperature-resistant components, such as blade disks, blades, and casings of aero-engines, and are key materials in a new generation of high thrust-to-weight ratio aero-engines. In the service environment of engineering applications, the creep resistance of high-temperature titanium alloys is one of the most important characteristic indicators. This paper reviews and analyzes the research status and progress on the creep properties of typical high-temperature titanium alloys in service in recent years. The effects of the creep parameters, alloy composition, and microstructure on the creep behavior of high-temperature titanium alloys are discussed, and various possible mechanisms for increasing the creep resistance of high-temperature titanium alloys are summarized.

1. Introduction

High-temperature titanium alloys refer to disordered, solid-solution-strengthened titanium alloys with a service temperature greater than or equal to 300 °C, and high-temperature titanium alloys are also known as heat-resistant titanium alloys [1,2]. High-temperature titanium alloys, with their excellent performance characteristics, can well meet the requirements of higher-performance aircraft engines [3]. As important structural materials for key hot-end components, such as compressor blades, blade disks, and engine casings, they significantly reduce the aircraft weight, improve the thermal strength and optimize the structural efficiency [4]. However, with the continuous development of advanced aircraft engines in terms of operating temperature, thrust-to-weight ratio, service life, and safety and reliability, the performance requirements of high-temperature titanium alloys are becoming increasingly demanding. They not only require the material to have instantaneous strength, from room temperature to high temperatures, but it also needs to have a good balance of a variety of properties [5,6], such as fatigue strength, creep strength, thermal stability, and high-temperature creep resistance. The revolutionary development of high-temperature titanium alloys is closely related to the progress of aircraft engine design. After half a century of unremitting efforts, the performance of traditional high-temperature titanium alloys has improved, from TC4 [7] and IMI318 [8] at a long-term service temperature of 300 °C to IMI834 [9] and Ti-1100 [10] at 600 °C applications, and the dosage in aircraft engines has gradually increased, as shown in Figure 1 [11]. At present, almost all of the high-temperature titanium alloys applied above 600 °C are Ti-Al-Sn-Zr-Mo-Si series near-α titanium alloys [12]. The performance of high-temperature titanium alloys has become the most important exponent to measure the advancement of aircraft engines.

During the service process, high-temperature titanium alloys are subjected to the long-term effect of internal and external mechanical loads, which is likely to lead to material failure, resulting in the size and performance of parts that cannot meet the requirements of use [13,14]. Especially when the service temperature exceeds 500 °C, the importance of the creep resistance and thermal stability of titanium alloys becomes increasingly prominent [15,16,17]. In general, thermal stability refers to the ability of high-temperature titanium alloys to maintain structural stability after long-term exposure to high temperatures. For high-temperature titanium alloys, on the one hand, the mechanism of microstructural instability is mainly affected by the dissolution, segregation and second phase precipitation caused by diffusion, and the increased diffusivity also leads to microstructural degradation by means of recrystallization and phase transition [18], but diffusion is an important mechanism of high-temperature creep, so that creep resistance and thermal stability maintain unity. On the other hand, in the process of high-temperature creep, the precipitation of the second phase in the grain boundary and in the grain boundary can hinder the sliding of dislocation and the grain boundary, which is conducive to reducing the creep rate, and these second phases in the grain boundary and in the grain boundary will hinder the coordination of plastic deformation between the grains, resulting in decreased thermal stability, so creep resistance and thermal stability remain contradictory. At an extreme service temperature, the contradiction is often particularly pronounced, which is also the key to limiting the use of high-temperature titanium alloys at higher temperatures. Therefore, it is very important to study how to balance the creep resistance and thermal stability of high-temperature titanium alloys. For high-temperature titanium alloys, the creep resistance in high-temperature service environments determines their service life and the safety and reliability of components, so the study of their creep resistance is of revolutionary significance for expanding their engineering applications.

Numerous research results show that there are many factors affecting the creep properties of high-temperature titanium alloys [19,20,21], including ambient temperature, applied stress, microstructure, phase size, precipitated phase content and distribution, and impurity element content, as shown in Figure 2. Therefore, this paper mainly focuses on the influence of the above aspects on the creep properties of high-temperature titanium alloys and the possible mechanisms for improving their creep resistance to provide reference for the preparation of high-temperature titanium alloys that can be used at higher temperatures and that have a good match between thermal stability and creep resistance.

2. Creep Research on High-Temperature Titanium Alloys

2.1. Creep Phenomenon and Law

Creep usually refers to the slow plastic deformation of a material under the action of high temperature and continuous loading. After applying a constant stress to the sample at a certain temperature, the curve of strain ε changing with time t, as shown in Figure 3, can be obtained, which is, namely, the typical creep curve under constant stress [22]. According to the shape of the creep curve, it is mainly divided into three stages: (Stage I) The strain is generated at the moment of loading, the creep rate gradually decreases over time, indicating that the creep resistance gradually increases, or the material is hardened; this stage is called the initial creep stage. (Stage II) The creep curve is similar to a straight line, and the creep rate remains unchanged; this stage is called the steady-state creep stage. (Stage III) The creep rate sharply increases over time until, finally, fracture occurs; this stage is called the accelerated creep stage. When the creep stress is small, the creep curve only displays Stage I and Stage II but not Stage III. When the creep stress is high, Stage II is very short, and the creep enters the third stage very quickly. When the creep temperature is very high and the stress is very large, the creep curve passes directly from Stage I to Stage III without going through Stage II. When the total creep strain reaches a certain value, creep fracture of the material can occur.

In general, the material will creep from low temperature to near the melting point, T_m, but the creep of engineering significance occurs mainly between 0.4 T_m and 0.7 T_m. The higher the melting point of a material, the better its creep resistance. The creep resistance of a material is generally expressed by the steady-state creep rate, $\dot{\mathsf{\epsilon }}$. The steady-state creep rate is not only related to the service temperature, T, and the applied stress, σ, but it is also closely related to the characteristic factors of the material itself, such as the grain size, microstructure, precipitate content and distribution, and impurity element content. To characterize the type of steady-state creep deformation behavior, the stress exponent n is usually used as an exponent to reflect the essential characteristics of steady-state creep, and different stress exponent ranges represent different creep mechanisms. The larger the stress exponent, the greater the influence of the stress on the creep deformation. According to a large number of test results, the relationship between $\dot{\epsilon }$ and T and σ is obtained according to the power law creep that is obtained, that is, the creep constitutive equation:

$$\dot{\epsilon }=\mathrm{A}{\mathsf{\sigma }}^{\mathit{n}}\exp \left(-\frac{{\mathit{Q}}_{\mathit{c}}}{\mathrm{R}\mathit{T}}\right)$$

(1)

where A represents the constant related to the temperature and material properties; Q_c is the apparent creep activation energy of the material; R is the gas constant, which has a value of 8.3145 J/(mol·K). This constitutive relationship is widely used in the study of the steady-state creep properties of titanium alloys.

2.2. Classical Creep Theory

There are many mechanisms of creep deformation, such as dislocation slip and climb, grain boundary slip, and vacancy diffusion, which can lead to creep plastic deformation. The contribution of different deformation modes to creep deformation varies with the change in temperature and stress. Generally, under conditions of different grain sizes, different temperatures and applied stresses, the dominant mechanism of steady-state creep behavior can be roughly divided into the Harper-Dorn (H-D) creep mechanism [23,24], diffusion creep mechanism [25,26], grain boundary slip creep mechanism [27,28], and dislocation slip or climb creep mechanism [29,30]. In order to analyze the influence of the grain size, temperature, and stress on the creep mechanism more clearly, Langdon and Mohamed et al. [31] constructed two types of creep mechanism diagrams, as shown in Figure 4. By adopting the rules summarized by Langdon and Mohamed, the critical values of the changes among the different creep mechanisms can be calculated [32]. These creep mechanisms have also been used to explain the creep of high-temperature titanium alloys under certain conditions. Several important creep mechanisms are described in detail below.

The creep mechanism based on the H-D mainly occurs in the case of high temperature, low stress, and coarse crystal. The basic feature is that the creep rate is proportional to the applied stress, and the stress exponent n = 1 [32]. The activation energy of the creep is equal to the activation energy of the lattice self-diffusion, indicating that the H-D creep is the result of the grain boundary dislocation climbing [33]. The H-D creep occurs only at the steady-state creep stage, and the creep rate is independent of the grain size. However, Blum [34,35] recently questioned the existence of H-D creep; they suggested that the early conclusions on H-D creep grinding may be due to insufficient accumulation of strain to achieve a steady state. In short, whether classical H-D creep exists is still a difficult problem that needs to be proved, and whether the stress exponent of the five-power law creep decreases to 1 with a reduction in the applied stress is also a key problem that needs to be solved urgently [36].

Diffusion creep generally occurs in a higher temperature environment of an approximately specific temperature (T/T_m) ≥ 0.5, and the diffusion creep curve does not appear in Stage I. According to classical creep theory [37,38,39], the diffusion-based creep mechanism has two main forms of deformation, one is Nabarro-Herring (N-H) creep and the other is Coble creep, and the creep rates are all closely related to the grain size. The N-H creep mechanism [40], also known as the volume diffusion mechanism, mainly occurs at a high temperature, low stress and medium grain size. The creep rate is inversely proportional to the square of the grain size. The creep activation energy is equal to the self-diffusion activation energy of the crystal lattice. When creep occurs at low temperature, low stress and fine grain, grain boundary diffusion takes precedence over lattice diffusion. This creep mechanism is known as Coble creep [41], also known as the grain boundary diffusion mechanism. The Coble creep rate is inversely proportional to the cube of grain size, and its activation energy is equivalent to the grain boundary diffusion activation energy, approximately half of the N-H creep activation energy. Under certain conditions, these two mechanisms can occur simultaneously, and the stress exponent is n = 1 [22].

The creep mechanism based on dislocation mainly occurs at higher temperatures and higher stresses. When the load is applied, the material will creep and deform, and the dislocation in the slip plane will produce dislocation multiplication during the slip process until it becomes immobile when it encounters obstacles (such as precipitates, solid solution atoms, and other dislocations), and this process is called work hardening (i.e., dislocation density increase). While at high temperatures, the energy diffusion of the atom is intensified by heat activation, and the ability of the dislocation motion (such as slip, climb, and cross slip) is enhanced, thus overcoming some short-range obstacles and causing the plastic deformation to continue. This process, which occurs under the action of thermal activation, is called recovery softening (i.e., dislocation density decrease), as shown in Figure 5 [29]. In general, steady-state creep is the result of the dynamic balance between work hardening and recovery softening [42]. The creep stress exponent, n, of a titanium alloy’s mechanism is dominated by dislocation motion, and it is usually distributed in the range of 3–5. In recent years, some researchers have also proposed an improved jogged screw dislocation model, which is considered the main mechanism that controls the steady-state creep mechanism of γ-TiAl alloy [43], Ti-6242Si alloy [44], and Zr-4 alloy [45] under specific stress conditions and ambient temperature, and its stress exponent, n, is generally between 4 and 6.

Grain boundary sliding plays an important role in the creep deformation of polycrystalline metallic materials at high temperature [46]. When the temperature is higher than 0.4 T_m, due to the easy diffusion of atoms at the grain boundary, the adjacent grains will slide relatively at their common interface by shear motion after the stress is applied, thus promoting the creep deformation [38,46]. Grain boundary sliding can sustain the creep deformation alone or as part of the creep mechanism, and the stress exponent, n, is about 2 [47]. When grain boundary sliding occurs as an independent creep mechanism, the steady-state creep rate and stress meet certain rules [48]. Under different stress conditions, the mechanism of coordinating grain boundary sliding is different, as shown in Figure 6 [22]. If the grain boundary sliding is not fully coordinated, the hole will form at the grain boundary. Under the condition of very low stress, grain boundary sliding is mainly coordinated by the elastic deformation within the grain. Under low stress conditions, diffusion creep is generally required to coordinate grain boundary sliding. In recent years, during the creep deformation of alloys, it is generally believed that the ratio of grain boundary sliding strain to total creep strain increases with increasing temperature, decreasing the stress and decreasing the grain size. However, in general, the proportion of grain boundary slip to the total creep strain is not high, generally about 10% [49].

For alloys strengthened by second phase precipitation, the steady creep rate is obviously lower than that of alloys without the dispersion phase precipitation, and the finer the grains and the smaller the spacing, the better the creep resistance. In addition, the stress exponent, n, is generally 7–8 or even 10–40. The creep activation energy is much greater than the self-diffusion activation energy of the matrix. Up to now, research has mainly focused on the mechanism of dislocation on the second phase particles. When the applied stress is higher than the Orowan stress, the dislocation bypasses the second phase particles through the Orowan mechanism, which is a nonthermally activated process and insensitive to temperature change [50]. When the applied stress is less than the Orowan stress, the dislocation passes through the second phase particles by climbing [51]. When the second phase particles are soft and coherent with the matrix, the dislocation continues to slip through the second phase particles by the cutting mechanism, and the shear stress required by the dislocation to cut the second phase particles is the creep threshold stress [52]. According to research results, the creep threshold stress is equal to half of the Orowan stress. If the applied stress is less than the threshold stress, the alloy will not undergo creep deformation [53]. In general, there are still many questions regarding the source of the creep threshold stress and the creep mechanism of second-phase precipitation strengthened alloys, and further studies are needed.

3. Factors Affecting Creep Behavior of High-Temperature Titanium Alloys

3.1. Creep Parameters

In the process of creep deformation, the creep rate and total creep strain of high-temperature titanium alloys increase with the increase in stress and temperature, and the total creep strain will gradually accumulate with the extension of time. Cho W, et al. [54] studied the creep behavior of the near-α Ti-6242 (Ti-6Al-2Sn-4Zr-2Mo) alloy at a 760 °C high temperature, indicating that the steady-state creep rate is exponentially related to the stress under high-stress conditions. The steady-state creep rate under low stress has a linear relationship with the stress. When Si was added to Ti-6242 alloy to form Ti-6242Si (Ti-6Al-2Sn-4Zr-2Mo-0.07Si) alloy to study the high-temperature creep properties, Es-Souni M, et al. [55] also reached the same conclusion. Viswanathan G B, et al. [44] compared the creep properties of Ti-6242 alloy at 565 °C. At the high stress (310 MPa), transmission electron microscopy (TEM) analysis showed that the creep deformation was dominated by the motion of the α-type 1/3<11$\stackrel{\mathrm{-}}{2}$0> screw dislocation, n = 1. At the lower stress (172 MPa), the creep deformation was dominated by the climbing α-type 1/3<11$\stackrel{\mathrm{-}}{2}$0> edge dislocation; the dislocation motion rate had no obvious effect on the creep deformation process, n = 5–7, and the creep mechanism was consistent with the Harper–Dorn creep model. Yang D Y, et al. [19] conducted high-temperature creep experiments on a new near-α high-temperature titanium alloy: Ti-6.6Al-4.6Sn-4.6Zr-0.9Nb-1.0Mo-0.32Si. The microstructure was quantitatively investigated using optical microscopy (OM), scanning electron microscopy, and TEM. It was shown that the steady-state creep rate of the alloy increases with an increase in the stress and temperature. The creep mechanism gradually changes from dislocation climbing to dislocation viscous slip. The steady-state creep rate of IMI834 (Ti-5.8Al-4Sn-3.5Zr-0.7Nb-0.5Mo-0.35Si-0.06C) alloy [56] at 873 K and 150 MPa is significantly lower than that of IMI834 (Ti-5.8Al-4Sn-3.5Zr-0.7Nb-0.35Si-0.06C). The main reason is that the Ti₃Al phase and grain boundary precipitated during the creep process can inhibit the dislocation movement and improve the creep resistance.

In general, different creep stresses and temperatures can lead to different creep mechanisms for the same alloy. In combination with relevant literature, for near-α titanium alloys, Bania P J, et al. [57] also studied the creep behavior of Ti-6242Si (Ti-6Al-2Sn-4Zr-2Mo-0.07Si) alloy and found that there are different creep mechanisms in different temperature ranges: In the range of 496–565 °C, the creep deformation process is controlled by high activation energy, obeying the power law constitutive equation, and the creep mechanism is dislocation climbing. In the range of 454–482 °C, the creep deformation process is controlled by low activation energy, and the power law fails. The creep mechanism is mainly dominated by dislocation cross slip or dislocation slip cutting forest dislocation. Zhao Y Q, et al. [58] demonstrated that Ti811 (Ti-8Al-1Mo-1V) alloy has different creep mechanisms at lower temperatures (325–425 °C) and higher temperatures (425–525 °C). The reasons for the change in creep mechanism was mainly analyzed with SEM and TEM techniques. At 325–425 °C, a large amount of the Ti₃Al phase precipitated in the alloy increases the creep resistance with the increasing temperature, and the creep strain increases slowly, indicating that the temperature has a great influence on the creep. When the temperature is 425–525 °C, the influence of the temperature on the creep is much less, because at high temperatures, the grain boundary slides to form holes, the hole initiation, and growth; the grain boundary dislocation aggregation produces stress concentration, and then causes microcracks and crack growth, finally leading to creep fracture.

For α + β titanium alloys, Li X X, et al. [59] also found, in their study on the high-temperature creep mechanism of TC6 alloy, that there are different creep mechanisms at different temperatures. The OM and TEM techniques were used to observe and analyze the microstructure of the alloy before and after creep. It is suggested that the alloy has a dual mechanism of dislocation slip and diffusion at 350–450 °C. The creep mechanism is diffusion creep at 450–500 °C, and the stress exponent, n, increases from 2 to 4.9 with the increase in temperature. Qi Y L [60] studied the creep behavior of TC11 alloy at high temperature. When the alloy creeps at 450–500 °C/300–450 MPa, the stress exponent, n, is 2.13, and the creep mechanism is mainly controlled by dislocation climbing. When the alloy creeps at 550 °C/300–350 MPa, the stress exponent, n, is 5.2, and the creep mechanism is mainly controlled by dislocation slip. When the alloy creeps at 550 °C/400–450 MPa, the stress exponent, n, is 1.5, and the creep mechanism is mainly controlled by dislocation climbing and diffusion. Badea L, et al. [17] also analyzed the law of steady-state creep rate variation with stress in Ti-6Al-4V alloy, which shows that in the range of 450–600 °C, when the stress is less than 0.9×YS, the steady-state creep strain rate variation with stress follows the power law creep, the stress exponent, n, is 4–5, and the creep mechanism is the diffusion mechanism. When the stress is greater than 0.9 × YS, the power law failure occurs, and the creep mechanism is dominated by the climbing process of dislocation. Zong Y Y, et al. [61] also reached the same conclusion in their study on Ti-6Al-4V alloy: creep deformation of the alloy at temperatures above and below 650 °C is controlled by vacancy diffusion and dislocation climbing, respectively.

For β titanium alloys, Wang M M, et al. [62] calculated the steady-state creep state, stress exponent, and creep activation energy of Ti40 alloy under different stresses and temperatures, and on this basis studied its creep mechanism. The test results showed that the creep activation energy of the alloy is lower than the self-diffusion activation energy of β-Ti at 460–540 °C and 250 MPa, which is comparable to the diffusion activation energy of other solid solution elements in Ti. Therefore, it is considered that the creep is controlled by the double mechanisms of dislocation climbing and diffusion, and it has good creep resistance. At the creep temperatures of 540–620 °C and 250 MPa, the dislocation motion resistance becomes smaller and the diffusion of atoms and vacancies further increases. It is believed that the creep is controlled by the diffusion mechanism, and the grain boundary movement also contributes to the creep, but the creep resistance is poor. Xin S W, et al. [63] also studied the creep properties of Ti40 alloy, and concluded that in the low-temperature range (500–520 °C), the stress concentration caused by the plug dislocation mainly depends on pipe diffusion and Si element diffusion (precipitation), and the climb occurs. Because of the low activation energy of this process, the dislocation involved in recovery is less, and the steady-state creep rate of the alloy is low. The stress has a great influence on the dislocation slip. Thermal activation controls the dislocation climbing (when the stress is released, the new dislocation can be generated) and controls the steady-state creep deformation. As the temperature increases, the effect of diffusion on creep deformation becomes more and more obvious. In the high-temperature region (535–550 °C), the degree of the dislocation packing decreases greatly and the dislocation climb associated with recovery increases sharply, resulting in a sharp increase in the steady-state creep rate and a significant reduction in steady-state creep time. The creep mechanism in this process is mainly controlled by self-diffusion or alloy element diffusion.

3.2. Alloy Composition

Pure titanium has excellent plasticity and thermal stability, but its tensile strength is low and its creep resistance is poor. In order to meet the requirements in actual production, some alloying elements must be added to form titanium alloys. At present, titanium alloys can be divided into five types of titanium alloys—α titanium alloy, near-α titanium alloy, α + β titanium alloy, near β titanium alloy, and β titanium alloy—according to the phase structure and the content of β stable elements in the metastable state. The high-temperature titanium alloys widely used in the aerospace field are mainly α + β and near-α titanium alloys. Serving in environments below 500 °C is mainly the high-aluminum equivalent of α + β titanium alloy, while the more promising working temperature of above 500 °C is mainly the near-α titanium alloy. As for β titanium alloys, they are rarely used as superalloys due to their structural instability and relatively high diffusion coefficient, which make them of limited use in high-temperature applications. However, because of the special application of burn-resistant titanium alloy, its working environment requires that this alloy be used at a higher temperature for a long time, so research on the thermal stability and creep resistance of burn-resistant titanium alloy is also an important direction.

Table 1. Relevant parameters of high-temperature titanium alloys commonly used in countries around the world. Date from refs. [64,66].

Country	Alloy Grade	Nominal Composition	Alloy Type	Service Temperature/°C	Year
U.S.A	Ti-64	Ti-6Al-4V	α + β	300	1954
	Ti-811	Ti-8Al-1Mo-1V	Near α	425	1961
	Ti-6246	Ti-6Al-2Sn-4Zr-6Mo	α + β	450	1966
	Ti-6242	Ti-6Al-2Sn-4Zr-2Mo	Near α	450	1967
	Ti-6242S	Ti-6Al-2Sn-4Zr-2Mo-0.1Si	Near α	520	1974
	Ti1100	Ti-6Al-2.8Sn-4Zr-0.4Mo-0.45Si	Near α	600	1988
U.K.	IMI550	Ti-4Al-2Sn-4Mo-0.5Si	α + β	425	1956
	IMI679	Ti-2Al-11Sn-5Zr-1Mo-0.2Si	Near α	450	1961
	IMI685	Ti-6Al-5Zr-0.5Mo-0.25Si	Near α	520	1969
	IMI829	Ti-5.5Al-3.5Sn-3Zr-0.3Mo-0.3Si-1Nb	Near α	580	1976
	IMI834	Ti-5.8Al-4Sn-3.5Zr-0.5Mo-0.35Si-0.7Nb-0.06C	Near α	600	1984
Russia	BT321	Ti-6.5Al-2.5Mo-0.3Si-0.5Fe-1.5Cr	α + β	400–450	1957
	BT8	Ti-6.5Al-3.5Mo-0.2Si	α + β	500	1958
	BT9	Ti-6.5Al-2Zr-3.5Mo-0.3Si	α + β	500–550	1958
	BT18	Ti-8Al-8Zr-0.6Mo-1Nb-0.22Si-0.15Fe	Near α	550–600	1963
	BT18Y	Ti-6.5Al-2.5Sn-4Zr-0.7Mo-1Nb-0.25Si	Near α	550–600	1963
	BT25	Ti-6.8Al-2Sn-1.7Zr-2Mo-1W-0.2Si	α + β	500–550	1971
	BT25Y	Ti-6.5Al-2Sn-4Zr-4Mo-1W-0.2Si	α + β	500–550	1971
	BT36	Ti-6.2Al-2Sn-3.6Zr-0.7Mo-5W-0.15Si	Near α	600	1992
China	TC4	Ti-6Al-4V	α + β	300–400	1965
	TC6	Ti-6Al-2.5Mo-0.3Si-0.5Fe-1.5Cr	α + β	450	1965
	TC9	Ti-6.5Al-2.5Sn-3.5Mo-0.3Si	α + β	500	1965
	TC11	Ti-6.5Al-1.5Zr-3.5Mo-0.3Si	α + β	500	1979
	Ti-53311S	Ti-5.5Al-3.5Sn-3Zr-1Mo-0.3Si-1Nb	Near α	550	1986
	Ti55	Ti-5Al-4Sn-2Zr-1Mo-0.25Si-1Nd	Near α	550	1986
	Ti60	Ti-5.8Al-4.8Sn-2Zr-1Mo-0.35Si-0.85Nd	Near α	600	1994
	Ti600	Ti-6Al-2.8Sn-4Zr-0.5Mo-0.4Si-0.1Y	Near α	600	1994

summarizes the chemical composition, alloy types, service temperature, and production year of high-temperature titanium alloys developed by different countries [64]. It can be found that the development of high-temperature titanium alloys is very difficult, and the average annual service temperature can be increased to 3.75 °C. In addition, with the increase in the service temperature, the more complex the alloy composition system, the higher the content of the alloying elements. However, a large amount of alloying elements will lead to the precipitation of the brittle phase, which will reduce the mechanical properties, especially the thermal stability of the material. In order to coordinate the contradiction between the oxidation resistance, thermal stability, and creep resistance of high-temperature titanium alloys, it is particularly important to rationally adjust the composition and content of the alloying elements. At present, the influence of the alloy composition on the creep resistance of high-temperature titanium alloys has been extensively studied. Table 2 shows a comparison of the mechanical properties of typical high-temperature titanium alloys at 600 °C at home and abroad [65]. It can be found that the comprehensive properties of domestic high-temperature titanium alloys are not inferior to those of foreign ones.

In recent years, the general direction of the development of high-temperature titanium alloys around the world has basically been the same, and from the perspective of the alloy system, Ti-Al-Sn-Zr-Mo-Si series α-+-β-type titanium alloys and near-α titanium alloys are still the main kinds. In addition to containing more than 6% Al and a certain amount of Sn and Zr, the content of β stable elements is also high, and the content of the β phase is greater at room temperature. The representative alloys are Ti64 [67], Ti17 [68], and Ti6246 [69] (U.S.A.); IMI550 [70] (U.K.); BT6, BT8M-1, and BT9 [71] (Russia); and TC4 [72] and TC11 [73] (China). The content of β stable elements in near-α high-temperature titanium alloys is less (<2 wt.%), close to its solid solubility in the α phase, which consists of the α phase and a small amount of the β phase (3–10 vol.%), as well as two precipitated phases: silicide and the α₂ phase. It is worth noting that, here, the α phase and β phase are no longer pure titanium with an HCP structure or BCC structure but the solid solution phase with this as the matrix. The near-α high-temperature titanium alloys take into account the high creep strength of α titanium alloys and the high-temperature instantaneous strength of α + β titanium alloys. The representative alloys are Ti6242S [74] and Ti1100 [75] (U.S.A.); IMI829 [76] and IMI834 [9] (U.K.); BT25 [77], BT18y [78], and BT36 [79] (Russia); and Ti55 [80] and Ti60 [81] (China). The main features of the burn-resistant titanium alloys is the very high Mo equivalent, and the content of β stable elements is as high as 40 wt.%. At present, there are two main types: Ti-V-Cr series and Ti-V-Cu series. The comprehensive properties of the Ti-V-Cr series are better than that of the Ti-V-Cu series, so there are many studies on the creep properties of Ti-V-Cr series burn-resistant titanium alloys, which are represented by Alloy C (Ti-35V-15Cr) (U.S.A.), Ti-25V-15Cr-2Al-0.2C (U.K.), and Ti40 (Ti-25V-15Cr-0.2Si) (China). In general, the developmental trend of high-temperature titanium alloys has been from solid solution strengthening to ordered strengthening, and the manufacturing process has developed from reduction (such as the forging process) to incremental (such as the additive manufacturing process). Researchers have demonstrated that high-temperature titanium alloys produced using additive manufacturing have a similar level of high-temperature creep resistance below 500 °C as alloys produced with traditional processes [82]. However, the inherent limitations of these new materials and processes cannot completely replace solid-solution-strengthened high-temperature titanium alloys produced using forging/casting processes in a high-temperature range above 500 °C, so the development of new technical routes is an important technical direction for future research on high-performance titanium alloys.

Table 2. Performance comparison of typical 600 °C high-temperature titanium alloys. Date from ref. [65].

Alloy	Tensile Property at Room Temperature				Tensile Property at Room Temperature after 600 °C/100 h Thermal Exposure				Creep Resistance at 600 °C/150 MPa/100 h	Microstructure
	UTS /MPa	YS /MPa	EI /%	RA /%	UTS /MPa	YS /MPa	EI /%	RA /%	εr /%
Ti1100 [83]	1094	1008	8.8	11	1119	1038	5.2	7.6	0.09	Lamellar
IMI834 [84,85,86]	1089	983	8.5	17.5	1129	1045	3.2	-	0.1	Bimodal
	1041	959	8.5	13	-	-	-	-	0.07	Lamellar
BT36 [79]	1100	1049	15.1	36.3	-	-	-	-	-	Equiaxed
	1142	1085	13.4	36.5	-	-	-	-	0.2	Bimodal
	1159	1054	9.7	18.7	-	-	-	-	0.146	Lamellar
Ti60 [87]	1060	1010	11	24	1120	1050	7	8	0.1	Bimodal
	1080	1020	9	15	1100	1040	4	7	0.076	Lamellar
Ti600 [88,89]	1070	965	15	24	1070	995	6	15	0.099	Bimodal
	1060	955	11	19	1090	1010	6	13	0.081	Lamellar

Table 2. Performance comparison of typical 600 °C high-temperature titanium alloys. Date from ref. [65].

Alloy	Tensile Property at Room Temperature				Tensile Property at Room Temperature after 600 °C/100 h Thermal Exposure				Creep Resistance at 600 °C/150 MPa/100 h	Microstructure
	UTS /MPa	YS /MPa	EI /%	RA /%	UTS /MPa	YS /MPa	EI /%	RA /%	εr /%
Ti1100 [83]	1094	1008	8.8	11	1119	1038	5.2	7.6	0.09	Lamellar
IMI834 [84,85,86]	1089	983	8.5	17.5	1129	1045	3.2	-	0.1	Bimodal
	1041	959	8.5	13	-	-	-	-	0.07	Lamellar
BT36 [79]	1100	1049	15.1	36.3	-	-	-	-	-	Equiaxed
	1142	1085	13.4	36.5	-	-	-	-	0.2	Bimodal
	1159	1054	9.7	18.7	-	-	-	-	0.146	Lamellar
Ti60 [87]	1060	1010	11	24	1120	1050	7	8	0.1	Bimodal
	1080	1020	9	15	1100	1040	4	7	0.076	Lamellar
Ti600 [88,89]	1070	965	15	24	1070	995	6	15	0.099	Bimodal
	1060	955	11	19	1090	1010	6	13	0.081	Lamellar

3.2.1. α Stable Element

Among the α stable elements, Al is the most important alloying element in high-temperature titanium alloys, and almost all high-temperature titanium alloys contain Al. According to the Ti-Al binary phase diagram, it can be seen that the solid solubility limit of Al in Ti is 8 wt.%, and when the content of Al is greater than 8 wt.%, the ordered Ti₃Al phase (α₂ phase) is easily precipitated after long-term aging at high temperature. In addition, if the content of Al is greater than 6 wt.% and the alloy contains Zr, Mo, V, or Si and other elements at the same time, the alloy will precipitate the α₂ phase after high-temperatures and a short aging period [90]. Therefore, in order to maximize the solid solution strengthening effect of Al and to avoid the embrittlement of the alloy caused by the addition of excessive Al, the Al content is generally controlled to be below 2.25–6 wt.% [91]. Traditionally, the range of the alloy composition can be designed according to the famous empirical formula of aluminum equivalent proposed by Rosenberg et al. [92]. The formula fully explains the precipitation tendency of the α₂ phase, but it fails to express the physical significance of the aluminum equivalent factor of each alloying element. Therefore, it has certain limitations in its use as a design standard for new alloying elements. On the basis of this, Li D et al. [93,94,95] proposed the electron concentration theory, which can be used to judge the thermal stability of alloys. The electron concentration theory can also be used to determine the equivalent factor of each alloying element, which can overcome the disadvantages of the empirical formula of aluminum equivalent. When the average electron concentration, $\stackrel{\mathrm{-}}{\mathit{N}}$_P, is less than 2.12, there is no precipitation of the α₂ phase in the matrix. Therefore, the expression $\sum {\stackrel{\mathrm{-}}{\mathit{N}}}_{\mathit{i}}{\mathit{f}}_{\mathit{i}}^{\mathit{a}}\le \stackrel{\mathrm{-}}{\mathit{N}}$_P can be used as the thermal stability basis for the non-precipitation of the α₂ phase in the alloy.

However, in order to ensure the high-temperature strength and sufficient solid solution strengthening effect of the alloy, the upper limit of the critical equivalent or a certain degree greater than the critical equivalent value is generally used to design the alloy. Therefore, the existing high-temperature titanium alloys will observe the α₂ phase during the long-term creep process at high temperatures. According to relevant studies [96], a large amount of the α₂ phase precipitation and growth will reduce the plasticity of the alloy, but the dispersed α₂ phase can reduce the creep deformation rate of the alloy at high temperature, and the more α₂ phase precipitation, the larger the size, and the more obvious the improvement in the creep resistance of the alloy. The influence of ordered α₂ on the creep property of the alloy is mainly due to the existence of α₂ phase which changes or affects the motion state of the dislocation in the alloy. When α₂ is introduced into high-temperature titanium alloys, α + α₂ two phase strengthening is formed. Specifically, the existence of the α₂ phase promotes the planar slip and localizes the slip during deformation, especially the prismatic plane slip, mainly because the antiphase boundary energy on the prismatic plane is lower than that on the basal plane [97]. At the same time, the occurrence of the deformation twins is inhibited [98]. In a study of Ti-6Al alloy, Neeraj T et al. [99] found that the precipitation of the α₂ phase promoted the dislocation to move forward in the form of dislocation pairs. The slip was concentrated in the long dislocation pile, and there was almost no dislocation activity outside the pile; the plane sliding mainly occurred in the basal plane, prismatic plane, and pyramidal plane, and the frequency of the cross slip and climb was low; the creep resistance was excellent. Rosenberger A H et al. [75] studied the influence of α₂ relative creep behavior in Ti1100 (Ti-6Al-2.8Sn-4Zr-0.4Mo-0.45Si) alloy and concluded that the creep resistance is optimal when the α₂ phase is completely precipitated, while the creep resistance is weak when the α₂ phase is partially precipitated. The creep resistance is the worst in the solid solution state. The creep experiment in this study has the advantage of generating a complete creep curve in a short amount of time, which can be used to determine the creep resistance of Ti1100 alloy at high temperature. Xin S W, et al. [100] also came to the same conclusion in their research on the relationship between the α₂ phase and creep properties in Ti-600 alloy. Therefore, introducing the α₂ phase into high-temperature titanium alloys and effectively controlling the size and content of the α₂ phase is an effective way to improve the comprehensive properties of the alloys.

The precipitation of the α₂ phase is closely related to the addition of other alloying elements. In addition to the elements that can promote the precipitation of the α₂ phase, some elements also inhibit the precipitation of the α₂ phase, such as C [101], Nb [102], W [103], Mo [104], and others. In a bimodal structure containing −15% primary α phase, the addition of C can reduce the size of the α₂ phase, decrease the volume fraction of the α₂ phase, delay the embrittlement caused by the α₂ phase, improve the dispersion strengthening effect of the α₂ phase, and improve the creep property of the bimodal structure. This effect is attributed to the fact that the addition of C reduces the concentration of Al in the αp phase and increases the distribution of Al and Mo in the β-transformed structure. The effect of C addition on the high-temperature creep resistance of titanium alloys has also been reported for other types of alloys. For example, Sun F S, et al. [105], in their study on the creep resistance of Ti-35V-15Cr-xC alloy, found that the addition of C promoted the nucleation of Ti₂C particles in the matrix and significantly improved the creep resistance of burn-resistant alloys.

The addition of a small amount of B is also beneficial for improving the creep resistance of alloy, because the addition of B can refine the grain [106], and the strong and hard TiB whiskers precipitated in the solidification process can prevent the movement of dislocations in the α phase and β phase [107]. Boehlert C J, et al. [108] showed that the creep resistance of as-cast Ti-6Al-4V-1B alloy was significantly better than that of Ti-6Al-4V-0.1B alloy, and the steady-state creep rate was, at least, one order of magnitude lower than that of as-cast Ti-6Al-4V alloy. Chen W, et al. [109] investigated the as-cast Ti-6Al-2Sn-4Zr-2Mo-0.1Si alloy and showed that the addition of B could also improve the high-temperature creep resistance of the alloy, but the improvement’s effect was not as obvious as that of Ti-6Al-4V alloy.

3.2.2. Neutral Element

Sn and Zr are neutral elements that have high solid solubility in α-Ti and β-Ti, and they are combined with other elements to play complementary and strengthening roles [110]. Sn can reduce the susceptibility of alloys to hydrogen embrittlement. The addition of Zr can refine the grain and inhibit the precipitation of the ω phase in the alloy. In addition, during the transformation of the β phase to the α phase, Si is precipitated from the lattice in the form of silicide [111]. Because of its similar size and valence to Ti, Zr can freely substitute for Ti, so the addition of Zr helps to promote the formation of (Ti, Zr)_x(Si)_y compounds. Moreover, with the increase in the Zr content, the degree of dissolution of the Si in the Ti matrix decreases, thus increasing the effective supersaturation for a given Si content [112]. Flower et al. [113] showed that the diffusion rate of Zr in Ti is much smaller than that of Si, and the nucleation growth rate of silicide is mainly dependent on Zr. In addition, compared with titanium alloys without Zr addition, the addition of Zr causes the precipitation of silicide to be more dispersed and uniform, effectively inhibiting the creep deformation of the alloy [114]. Narayana P L, et al. [115] showed that the precipitation of the α₂ phase and silicide can be effectively reduced by adding Hf instead of Zr. After tensile tests at room temperature and 650 °C, the thermal stability of the alloy was significantly improved compared with existing high-temperature titanium alloys. When Sn, Zr, and Si are added to high-temperature titanium alloys at the same time, the Ti in silicide is easily replaced by Sn, and the (Ti, Zr)_x(Si, Sn)_y compound finally forms. However, an ordered Ti₃Sn phase is formed when the Sn content in high-temperature titanium alloys is too high. Generally speaking, the α/α₂ solid solubility limit of the Ti-Sn system is 16.9 wt.% [116], and the precipitation of the Ti₃Sn phase is beneficial to the improvement of the creep resistance and high-temperature instantaneous strength; its main weakness is that the thermal stability of the alloy decreases.

3.2.3. Isomorphic β Stable Element

The precipitation of the Ti₃X (X = Al, Sn, Ga, et al.) phase is also affected by the synergistic effects of other elements. Among the β stable elements, V belongs to the β-Ti isomorphic element and can be used as an enhancer and stabilizer. As a strengthening agent, V is mainly reflected in the infinite solid solution in β-Ti, and it also has a certain solid solubility in α-Ti; the solid solution strengthening effect is significant. As a stabilizer, V can reduce the β-transition temperature and increase the β phase stability. Therefore, the addition of V is conducive to improving the creep properties of titanium alloys. Ti-6Al-4V alloy exhibits a low strain rate and good creep resistance during creep [117]. The alloying feature is that both the α phase stable element Al and β phase stable element V are added to strengthen the microstructure of the α phase and β phase at the same time. In addition, V can also inhibit the precipitation of the Ti₃Al phase and avoid the phenomenon of alloy diffusion intensification during the formation of the Ti₃Al phase. Diffusion is an important mechanism of creep deformation, and an increase in diffusion will lead to an increase in creep deformation. Xin S W, et al. [118] prepared three kinds of burn-resistant titanium alloys with different V contents, tested their creep properties under different creep processes, and observed their microstructure after creep using SEM and TEM. The results show that there is an optimum content (25 wt.%) of V for the effect of V on the creep resistance of Ti-V-Cr burn-resistant titanium alloys. When the value is less than this value, with the increase in V content, the β phase will gradually stabilize, the vacancy formation position will gradually decrease, and V and Cr will more easily form an infinite solid solution, which hinders the nucleation of the TiCr₂ phase. The creep property of the alloy is gradually improved by decreasing the rate of Cr forming of segregation zone by diffusion. When the value is greater than this value, the β phase maintains the original stability, and the change in the V content has little effect on the creep resistance.

Mo, itself, does not form precipitates and plays a major role in solid solution strengthening, refining grains, and improving the hot working properties of alloys. A small amount of Mo (generally < 1 wt.%) is added to high-temperature titanium alloys to provide them. with a good balance between thermal stability and creep resistance. Zhao Y Q, et al. [119] studied the creep properties of Ti811 (Ti-8Al-1Mo-1V) alloy and concluded that the higher the Al content, the greater the dislocation motion resistance and the better the creep properties. When the content of Al is between 7.35 wt.% and 8 wt.%, the solid solution strengthening effect of Al is gradually enhanced. In this case, there is no Ti₃Al precipitation, and the inhibiting effect of Mo and V on the Ti₃Al precipitation cannot be reflected. Therefore, the change in the Mo and V contents has little influence on the thermal stability. However, the increase in the Mo and V contents can strengthen the β phase and hinder the movement of the dislocation between grains, so it is beneficial to improve the creep resistance. When the content of the Al is between 8 wt.% and 8.35 wt.%, the precipitation strengthening effect of Al is gradually enhanced, and the matrix is strengthened to the maximum extent. The influence of Mo and V can be masked, and there is no obvious effect on the dislocation movement, so the creep resistance is not affected. Because of the precipitation of the Ti₃Al phase in the alloy, the increase in the Mo and V contents can inhibit the precipitation of the Ti₃Al phase to a certain extent, and the plasticity of the thermal stability is significantly improved. At the same time, the toughness and the high-temperature tensile strength of the alloy are also significantly increased.

3.2.4. Eutectoid β Stable Element

Fe has an unusually fast diffusion behavior. The diffusion coefficient of Fe is 7 to 8 orders of magnitude faster than α-Ti, and 2 orders of magnitude faster than β-Ti. The addition of Fe to titanium alloys will increase the vacancy diffusion rate, improve the diffusion rate of α-Ti and other atoms, and then promote the creep process controlled by lattice diffusion [120]. Therefore, the content of Fe in titanium alloys designed for high-temperature applications is strictly limited. Huang S S, et al. [121] studied the effect of 0.05 wt.% and 0.2 wt.% Fe on the creep resistance of TA15 (Ti-6.5Al-2Zr-1Mo-1V) alloy at room temperature −500 °C. It is found that in the range of room temperature −200 °C, the increase in the Fe content enhances the solid solution strengthening of the β phase, which improves the creep resistance of the alloy. However, when the temperature is 300 °C and above, the presence of high amounts of Fe content will significantly deteriorate the creep property of the alloy, which is due to the enhancement of the Fe-induced diffusion effect. For the β stable element Ni, Hayes R W, et al. [122] studied the high-temperature creep behavior of trace Ni for Ti-6Al-2Sn-4Zr-2Mo alloy using the same rule. In the temperature range of 510–565 °C, Ni accelerates self-diffusion, and the creep resistance of alloy with high Ni content is worse than that of alloy with low Ni content.

V and Cr are widely used in the design of burn-resistant titanium alloys. Xin S W, et al. [123] believe that the increase in V and Cr contents is beneficial to the improvement of creep properties, the eutectoid reaction temperature of Cr-containing titanium alloy is 670 °C, and the instability of Cr in β-Ti allows for the easy formation of the TiCr₂ phase during long-term aging or thermal exposure; that is, the segregation region of Cr is formed first. Then, it is further transformed into the TiCr₂ phase by ordering. In addition, Cr is stronger than V due to the difference in their atomic radius (Ti has an atomic radius of 0.145 nm; the atomic radius of Cr is 0.127 nm; and V has an atomic radius of 0.135 nm) caused by the difference in the distorted stress field in the alloy. In a study of burn-resistant titanium alloy, Zhao Y Q, et al. [124] concluded that the optimum Cr content was 15%, because when the Cr content was low, the burn-resistant property of the alloy could not be satisfied, and the strength at room temperature was poor. When the content of Cr was high, the creep property was not affected, but the plasticity and thermal stability at room temperature could not be guaranteed.

For alloying elements with a high melting point, such as Nb, Ta, and W, the addition method and amount of their addition to high-temperature titanium alloys should be given great attention, especially for W, which has a melting point above 3400 °C. High melting point inclusion is a metallurgical defect that must be strictly controlled. W is also a eutectoid β stable element, but the eutectoid reaction temperature of W is higher at 715 °C, so high-temperature titanium alloys containing W have better thermal stability than high-temperature titanium alloys containing Cr. In addition, the higher the melting point of the solid solution element, the larger the radius difference between the element and the matrix atom, and the better the high-temperature creep resistance of the alloy. At present, the W-containing high-temperature titanium alloys that have been studied at home and abroad mainly include BT36 [79], TC25G [125], and Ti65 [126]. Solving the problem of adding W element is of great significance for further breaking through the “thermal barrier” temperature of 600 °C.

As an interstitial eutectoid element, Si has not received much attention. It was not until the 1970s, when Seagle R, et al. [110] first found that the addition of Si had a unique effect on the high-temperature creep resistance of Ti-6242 (Ti-6Al-2Sn-4Zr-2Mo) alloy, that the importance of Si received widespread attention. In a study of Ti-Si binary alloy, or multicomponent titanium alloy [127], the researchers found that the creep deformation and steady-state creep rate of alloy containing Si are always much lower than that of alloy without Si. At present, almost all high-temperature titanium alloys used in environments above 500 °C are combined with 0.1–0.5 wt.% Si, such as Ti-6242S (Ti-6Al-2Sn-4Zr-2Mo-0.1Si), IMI834, Ti1100, and Ti60. As shown in Figure 7 [128], there are two forms of Si in titanium alloys: One is solid solution silicon in the matrix, which mainly plays a role in solid solution strengthening. Si has some solid solubility in both the α and β phases. The maximum solubility of Si in β-Ti is 3%, which occurs mainly at 1330 °C. The maximum solid solubility of Si in α-Ti is 0.45%, which occurs at 860 °C [113]. The atomic size difference between Si and Ti is large, the Si in the solid solution state in the matrix is easy agglomerated at the dislocation with heat treatment, a Cottrell atmosphere forms, and the interaction between Si and other solid solution atoms can effectively prevent the climbing of the dislocation and improve the creep resistance of the alloy. The other is the silicide state, which mainly plays the role of precipitation strengthening. When Si reaches a certain solution limit in the matrix, fine and dispersed silicide is precipitated, which will hinder the slip of the dislocation and grain boundary and also achieve the purpose of improving the creep resistance of titanium alloy. However, although the creep property of the alloy will not change to a great degree if the silicide is precipitated too much, it will adversely affect the ductility of the alloy at room temperature and after creep. Therefore, Si has an optimum level of content, which may not be reflected in the creep resistance but in the comprehensive mechanical properties. When the content is too high or too low, the best strengthening effect cannot be achieved [12,110]. Fentiman W P, et al. [129] demonstrated that the addition of 0.25 wt.% Si significantly improved the creep resistance and strength of Ti-679 alloy. However, Anti M L, et al. [130] proposed that the addition of less than 0.1 wt.% Si to Ti6242 alloy can significantly improve the creep resistance. Using OM and SEM, it was observed that a precipitated phase larger than 150 nm was not uniformly distributed, while a smaller precipitated phase of 20–100 nm was uniformly distributed. All of the silicide was predominantly situated next to the β phase in the alloy, either at the prior β grain boundaries or the β phase in between the α colonies. Singh A K, et al. [131] showed that the optimum content of Si in Ti-6Al-3Sn-3Zr alloy is 0.3 wt.%, and that the creep residual strain increases again after exceeding this value. In addition, Paton N E, et al. [128] studied the creep properties of Ti-Si alloy at high temperature and found that the creep properties of the alloy were closely related to the present state of Si by analyzing the creep rate data and TEM results. The more silicide precipitated, the worse the creep properties of the alloy, and the best creep properties could be produced when the Si element was completely dissolved in the matrix. For a long time, people have had different opinions on the mechanism of action of Si in improving the creep resistance of high-temperature titanium alloy [132], especially for the high-temperature titanium alloys used for temperatures above 600 °C, and the role of silicide has not been clearly determined, so in the future, in high-temperature creep research on titanium alloy, this should be specific to the alloy.

In the study of the simultaneous precipitation of silicide and Ti₃Al in high-temperature titanium alloys, the dominant role and mechanism of the two precipitates in creep deformation are still unclear. Zhao L, et al. [133], in their study on the creep properties of Ti60 (Ti-5.8Al-4Sn-3.5Zr-0.4Mo-0.4Nb-1.0Ta-0.4Si-0.06C) alloy, found that the precipitation of silicide can effectively pin the phase boundary, and the precipitation of the α₂ phase can hinder the dislocation slip and climb within the α lamella. Therefore, when a large amount of silicide precipitates out at the same time as α₂, both the α layer interface and the α phase matrix are strengthened, resulting in the smallest creep deformation of the alloy. When only silicide precipitates, but the α₂ phase does not precipitate, the content of solid solution silicon is relatively reduced because of the silicide precipitates, the dislocation movement in the matrix is less hindered, and the creep deformation in the phase is larger. Therefore, Ti60 alloy can obtain the best creep properties only when it is strengthened both in the phase and at the phase boundary. Therefore, it is necessary to strictly control the precipitation amount and size of the silicide and the α₂ phase, and it is considered that the relative change in the Si content in the solid solution caused by the silicide precipitation has little effect on the creep properties of the alloy. In general, the different effects of Si and silicide in the solid solution under creep conditions can be explained by the different deformation mechanisms of the materials under different applied stress levels.

Previously, Rosenberger A H, et al. [75] summarized the law of influence of two precipitates on the creep properties of Ti1100 alloy and analyzed it with a schematic diagram, as shown in Figure 8. From the diagram, it can be found that with the extension of the aging time, the increase in the α₂ phase content is beneficial to the improvement of the creep resistance of the alloy. The creep resistance of the alloy is weakened by the decrease in the Si content in the solid solution. In general, aging has little effect on the creep properties of Ti-1100 alloy. Xin S W, et al. [100] concluded different conduct with previous researchers in their study about the creep properties of Ti600 alloy after different heat treatments. They believe that α₂ phase has strong creep strengthening effect, and both solid solution Si and silicide have creep strengthening effect, but the strengthening effect is obviously weakened when solid solution Si forms silicide during dynamic precipitation. This is because the dynamic precipitation strengthening effect in the creep process is accompanied by the diffusion effect in the process of nucleation and the growth of the silicide, and the diffusion effect offsets the dynamic precipitation strengthening effect of the silicide. Therefore, the best creep properties of the alloy can be obtained only when the silicide and α₂ phase are completely precipitated after full aging. Yue K, et al. [126] studied the creep deformation behavior and mechanism of Ti65 (Ti-5.9Al-4Sn-3.5Zr-0.3Mo-0.3Nb-2Ta-0.4Si-1W-0.05C) alloy and concluded that the primary creep stage was mainly dominated by the strengthening mechanism of the interaction between the α₂ phase and dislocation. The steady-state creep stage is mainly a strengthening mechanism of the uniformly precipitated silicide along the α/β phase boundary to hinder the dislocation movement and limit the grain boundary sliding. In general, with the increasing in the service temperature of high-temperature titanium alloys, the presence of silicide and the α₂ phase plays an increasingly important role in improving the creep properties of alloys and dynamic precipitation strengthening during service. Therefore, it is of great significance to study the size, quantity, morphology, and compositional state of the silicide and α₂ phases in alloys.

Compared with Ti60 alloy, an increase in the content of the weak β stable element Ta and the addition of the high melting point element W [103] in Ti65 alloy not only effectively compensates for the loss of alloy strength caused by the decrease in the content of the β eutectic elements Mo and Nb but also improves the creep and durability properties of the alloy. In addition, the addition of trace C expands the processing window of the two-phase region, improves the relative temperature sensitivity of the primary α during heat treatment in the upper part of the two-phase region, promotes the uniform distribution of the Al in the α phase and β phase, and decreases the precipitation of the Ti₃X phase. At the same time, the solid solubility of Si increases, which effectively reduces the number and size of the silicide precipitates, and after high-temperature treatment, the distribution of the silicide is more uniform, which is more conducive to achieving the best match between the thermal stability and creep resistance of the alloy. At present, the long-term service temperature of Ti65 alloy is up to 650 °C, and the short-term service temperature is up to 750 °C [134].

3.2.5. Rare Earth Elements

The most commonly added rare earth elements in high-temperature titanium alloys are neodymium (Nd) [80], yttrium (Y) [135], gadolinium (Gd), scandium (Sc) [136], cerium (Ce) [137], erbium (Er) [138], etc. The existence of rare earth elements in titanium alloy is mainly of the following three types: first, as a small amount of rare earth soluble in the titanium matrix; second, in the form of rare earth oxides, such as Nd₂O₃, Y₂O₃, Gd₂O₃, Sc₂O₃, CeO₂, and Er₂O₃; third, in the form of intermetallic compounds or other complex compounds. Relevant studies have shown that rare earth elements play an important role in improving the high-temperature mechanical properties of titanium alloys at high temperature, which is mainly reflected in the following: ① the formation of noncoherent and dispersive rare earth oxides with the matrix through internal oxidation—rare earth oxides can not only reduce the oxygen content in the matrix but also form dislocation substructures and reduce the density of the mobile dislocation; ② the promotion of the migration of Al and Sn atoms in the matrix to rare earth oxides and inhibition of the precipitation and growth of Ti₃X phase, thus improving the thermal stability of the alloy; ③ the promotion of the uniform precipitation of fine silicide, and they hinder the dislocation movement and strengthen the alloy matrix; and ④ refinement structure: the refinement effect of different rare earth elements is different. At present, it has been proved that the addition of Sc can refine the structure and improve the strength and microhardness of the alloys [139]. However, there are few studies on the effect of Sc on the creep properties of titanium alloys at high temperature, especially the effect on the Ti₃X phase and silicates precipitation in titanium alloys, and the creep mechanism. Therefore, the creep resistance mechanism of Sc is an important research direction for the future.

At present, the typical titanium alloys with rare earth elements are Ti55 alloy, Ti600 alloy, Ti633G alloy, etc. The addition of different rare earth elements has improved the microstructure and properties of alloys to different degrees. Ti55 alloy (Ti-5Al-4Sn-2Zr-1Mo-0.25Si-1Nd) [140,141,142] is a Ti-Al-Sn-Zr-Mo-Si alloy with 1 wt.% Nd. The second phase particles composed of SnO and Nd₃Sn/Nd₅Sn₄, which are noncoherent and uniformly distributed within the matrix, are formed through oxidation in the melting process. This microstructure can significantly improve the thermal stability, fatigue strength, and creep resistance of the alloy below 550 °C. In Ti600 alloy (Ti-6Al-2.8Sn-4Zr-0.5Mo-0.4Si-0.1Y) [143], in which is added 0.1 wt.% Y, a small amount of Y exists in the solid solution, and the most of the remaining Y is precipitated in the form of the rare earth oxide Y₂O₃; both of these cause significant hindrance to the movement of the dislocation. In addition, Y can not only refine the structure but also reduce the aluminum equivalent and inhibit the precipitation and growth of the brittle α₂ phase. Furthermore, Y can also change the size and distribution of (Ti, Zr)₆Si₃ and the configuration of dislocation in the alloy. Therefore, the thermal stability and creep resistance of the alloy can be further improved. Ti633G alloy (Ti-6.5Al-3Sn-3Zr-0.3Mo-1Nb-0.2Gd) [144] is formed by adding 0.2 wt.% Gd based on the IMI829 alloy’s composition, and the temperature can reach 550 °C. The addition of Gd improves the dispersion of the α₂ phase and silicide in the matrix, which not only allows for the full precipitation strengthening effect of the second phase but also avoids the brittleness damage caused by the coarse particles of the second phase, such that an excellent match between the thermal stability and creep resistance is obtained.

3.3. Microstructure

According to early studies, the creep properties of high-temperature titanium alloys largely depend on their microstructure, and alloys with different microstructures have different creep mechanisms, thus corresponding to different creep deformations and creep lives [7]. In addition, microstructural adjustment is the best way to achieve a good match between the thermal stability and creep properties of high-temperature titanium alloys. The four basic microstructures of titanium alloy are mainly an equiaxed structure, bimodal structure, basketweave structure, and Widmanstatten structure. Different microstructures can be obtained using different heat treatment processes [145]. Lamellar structures (e.g., basketweave structure and Widmanstatten structure) have the best creep properties at high temperatures [146]. One reason is that the grain size of all lamellar structures is large, the grain boundary volume fraction is reduced, the interfacial energy is smaller, and the interfacial energy is slow. Another reason is that at high temperature, the strength of both the intragranular and grain boundary is lower than that at room temperature: the strength of the grain boundary decreases rapidly at high temperature, and the intragranular strength plays a major role in the strengthening. In a lamellar structure, each lamellar structure engages with each other, its binding strength is higher than that of other structures, and the relative sliding trend is weakened, so the creep resistance at high temperature is relatively good. In addition, when the layers in the lamellar structure are finer, the phase interface is greater, the sliding between the lamellar layers is larger, and the creep resistance begins to deteriorate. The equiaxed structure has the worst creep properties [129], and the bimodal structure falls between the equiaxed and lamellar structures [147]. This is because the β-transition structure around the equiaxed α phase provides a rapid diffusion channel for the deformation of the equiaxed α phase, making the equiaxed α phase more susceptible to slip deformation than the lamellar α phase. With the decrease in the volume fraction of the equiaxed α phase, the noncoherent α/β phase interface volume fraction decreases, and the lamellar α phase and β phase transformations in the β-transition structure are longer, making the grain boundary sliding difficult. Therefore, the creep resistance of the alloy gradually improves. For example, Ti1100 alloy is characterized by a Widmanstatten structure that is formed by forging in the β phase zone or solid solution treatment, which provides the alloy with excellent creep resistance [148]. In addition, Balasundar I, et al. [56] used an artificial neural network to calculate the effect of changes with individual microstructural features on the creep strain of the alloy, as summarized in Figure 9.

Although the basketweave structure and the Widmanstatten structure have the best creep properties, in actual production and application, because of the poor plasticity of the lamellar structure, the thermal exposure of the precipitates generated at the grain boundary easily causes plastic fracture of the alloy, and the thermal stability is impaired. Therefore, the lamellar structure is not selected in the service state of most alloys, and the bimodal structure containing a certain content of equiaxed α phase is generally used as the final state of the structures [149]. This is because the thermal stability and fatigue properties of the bimodal structure are relatively better, and for high-temperature titanium alloys used in aircraft engines, in addition to considering the high-temperature creep properties of the alloy, it is more important to improve the fatigue properties and thermal stability properties, among other factors. In near-α alloys with an equiaxed α phase content of more than 50%, such as Ti6242 [150], crack nucleation is likely to occur within the equiaxed α phase due to the high-density accumulation of grain boundary dislocations in the alloy’s structure when high-temperature creep properties are tested. However, when the volume fraction of the equiaxed α phase is reduced, the creep resistance and fracture toughness of the alloy are improved, and the plasticity and fatigue properties also decrease. In addition, Balasundar I [56], et al. studied the correlation between the microstructural and creep properties of near-α IMI834 alloy and concluded that at any cooling rate, increasing the solution temperature can obtain higher creep resistance and tensile strength but at the cost of plasticity and crack propagation resistance. This is due to the decrease in the volume fraction of the noncoherent interface of the α phase and β phase and the increase in the size of the β-transformed structure. Russo P, et al. [151] pointed out that the content of primary α phase has a great influence on the creep properties of Ti-6242Si at 510 °C but has little effect on creep properties at 427 °C. In addition, under the same creep condition, the steady creep rate of the complete β-transformed structure obtained by heat treatment in the β single-phase region is significantly lower than that of the bimodal structure obtained by heat treatment in the α + β two-phase region, but the low-cycle fatigue performance of the complete β-transformed structure is worse than that of the bimodal structure. Therefore, according to a large number of research results, when considering the best match between thermal stability and creep properties in high-temperature titanium alloys’ microstructures, the bimodal structure with primary α phase content of 10–20% is preferred, and heat treatment in the α + β phase region is an effective means to control the volume fraction of the equiaxed α phase [152,153,154].

At the same time, Xin S W [155] proposed a near-β dual-solution treatment process that can strictly control the mixed microstructure of titanium alloys with 15–20% primary α phase, 10–15% secondary lamellar α phase, and 65–75% third α phase + aging β phase, and they named the microstructure of these alloys a triple-modal structure, as shown in Figure 10. At present, many researchers have studied the creep properties of titanium alloys with a triple-modal structure. For example, Hosseini R, et al. [156] compared Ti6242S (Ti-5.7Al-1.9Sn-3.9Zr-2.0Mo-0.08Si), a near-α high-temperature titanium alloy with a triple-modal structure, with Ti6242S with conventional and bimodal structures. The results show that the yield strength and plasticity toughness of the triple-modal structure were significantly higher than those of the Widmanstatten structure and slightly lower than that of the bimodal structure. Under the creep condition of 540 °C/540 MPa, the creep rate of the triple-modal structure was significantly lower than that of the bimodal structure, and slightly higher than that of the basketweave structure. Therefore, it is proved that the comprehensive mechanical properties of the triple-modal structure are the most excellent. In addition, Li J X [157], et al. also confirmed this in a comparative study of the creep properties of the triple-modal structure with the Widmanstatten structure and bimodal structure. In addition, Grabovetskaya G P et al. [158] showed that the increase in the creep resistance of ultrafine-grained Ti-Al-V-Mo alloy obtained via severe plastic deformation after annealing is associated with the transformation of the β phase into the α phase and the redistribution of alloying elements, which increases the strength of the β phase and the appearance of grains with a lamellar structure in the alloy. This is due to the nonequilibrium temperature phase transitions and the formation of supersaturated solid solution in the process of forming the ultrafine-grained state via severe plastic deformation [159]. The ultrafine-grained structure with a nonequilibrium phase at high temperatures can not only recover the deformed structure but also lead to the decomposition and reverse phase transformation of supersaturated solid solution. Therefore, the effect of pre-crystallization annealing on the creep properties of ultrafine-grained alloys has a unique advantage compared to pure metals [160]. In general, the optimal heat treatment process is selected according to the actual service environment of the alloy, and the excellent mechanical properties of the alloy are provided by adjusting the microstructure.

4. Conclusions

The melting point of pure titanium is 1682 °C, which is 227 °C higher than the melting point of nickel. At present, the maximum operating temperature of nickel-based high-temperature alloys is 900–1000 °C, while the highest service temperature of titanium alloys is only 600 °C, and most of the high-temperature titanium alloys above 600 °C are the results of laboratory or semi-industrial trial productions. Therefore, in terms of the melting point, there is great development potential for the design and research of titanium alloys for long-term service at higher ambient temperatures. In recent years, many researchers have studied titanium alloys at higher service temperatures. However, it is still difficult to achieve a balance between a high-temperature creep property and thermal stability in order to obtain excellent comprehensive mechanical properties at high temperature. Therefore, it is a great challenge to develop high-temperature titanium alloys with excellent comprehensive mechanical properties at higher operating temperatures [161].

In general, for research on new high-temperature titanium alloys, we can carry out future studies starting from the following points: (1) Adding new trace elements—For example, α stable elements, such as Ga, In, Pb, and Sb, not only have an excellent solid solution strengthening effect in α-Ti but also have a much smaller tendency to form Ti₃X than Al and Sn, and they can refine the grain to a certain extent, so it can significantly reduce organizational instability. In addition, a β stable element, such as Bi, can be added to effectively improve the creep property of the alloy by forming Ti₂Bi. In short, the optimal design and accurate calculation of the composition and content of the added elements are expected to achieve the best combination. (2) The microstructural evolution is closely related to the mechanical properties. The optimum comprehensive properties of high-temperature titanium alloys can be achieved by reasonable technology for the heat working process and heat treatment regime. For example, through the near isothermal forging process in the α + β two-phase region and the near-β double solution + aging treatment process, the volume fraction of the primary α phase is controlled to be between 10 and 30%, the secondary α phase is between 10 and 15%, and the precipitation and distribution of the α₂ phase and silicide at the microscale are controlled so that the obtained triple-modal structure has a better match between the thermal stability and creep resistance. (3) Adding rare earth elements, such as Nd, Er, Gd, Y, Ce, and La—A large number of research results show that rare earth elements not only have high solubility, themselves, in titanium alloys but are also conductive to improving the solid solution strengthening effect of other elements. In addition, the rare earth oxides formed by internal oxidation mainly play roles in deoxidation, structure refinement, dispersion strengthening, reduction of stacking fault energy, and increasing the dislocation density. They play a beneficial role in improving the creep resistance and thermal stability of the alloy. (4) Research on high-temperature oxidation resistance coating of titanium alloys. Al coating can form a dense Al₂O₃ oxide film, inhibit the diffusion of oxygen elements into the matrix, and reduce the surface oxidation of titanium alloys. Pt and Au coatings can not only form a diffusion barrier of oxygen but also have a strong bonding force with the matrix under stress, and they do not crack easy. For example, two different types of surface modification techniques have been applied to coat the titanium alloys: electrodeposition and pack aluminizing. These result in an aluminum-platinum coating, which is a prospective coating material for preventing α-case formation and protecting against oxidation in components made of titanium alloys. Research on the structural design and coating technology of high-temperature oxidation-resistant coating is a key point for the long-term working temperature of titanium alloys to exceed 650 °C. (5) TiAl alloys based on intermetallic compounds, such as Ti₃Al, TiAl, and Ti₂AlNb, are being developed. The creep property of TiAl alloy is very sensitive to the structure type, and the full lamellar structure has the best creep resistance. The smaller the lamellar layer, the better the creep resistance.

This paper reviewed the research status on creep resistance, pointed out the development bottleneck of high-temperature titanium alloys, and provided several possible solutions, and it is expected to serve as a reference for the research and development of high-temperature titanium alloys.