Effect of Heat Treatment on the Microstructure and Mechanical Properties of High-Strength Ti–6Al–4V–5Fe Alloy
2019 年 60 巻 2 号 p. 269-276
詳細
2019 年 60 巻 2 号 p. 269-276
The effect of heat treatment on the microstructure characteristics and mechanical properties of the high-strength and low-cost Ti–6Al–4V–5Fe alloy was investigated. Two-phase (α + β) and single-phase (β) solution treatments and aging were applied to determine the relationship between microstructures and properties. The size of the grain after treatment of the solution in the β single phase was only dozens of micron for the primary α(αp), which exhibited an obvious pinning effect on grain growth. The morphology and volume fraction of αp phase were highly sensitive to the heat treatment temperature and remarkably influenced the properties of the alloy. When the solution temperature was 780°C and the aging temperature was 550°C, the largest proportion (40%) of the globular αp phase and small secondary α(αs) phases resulted in the best performance of the alloy, with an ultimate strength of up to 1300 MPa and 9.57% elongation. The fracture surface of tensile specimens was systematically studied, showing a ductile mode of tensile failure after the sample was treated below 800°C and then aged. However, it exhibited a brittle mode when the alloy is treated above 800°C.
Fig. 8 Results of the elemental mapping of the surface of alloy samples that were heated at different temperatures. (a) (e) and (i) was solution treated at 700°C, 780°C and 820°C, respectively, then aged at 500°C. (a) (e) and (i) SEM images. (b)–(d), (f)–(h) and (j)–(l) EPMA elemental maps of Fe, V and Al, respectively.
In recent years, titanium alloys have been used successfully in many fields, such as automotive parts, down-hole service, surgical implant and aerospace industry, due to their high strength-to-density ratio, excellent biocompatibility, good corrosion resistance, excellent fatigue performance, good environmental resistance and rich microstructural features.1–9) The Ti–6Al–4V alloy developed by the United States in 1954 is the most widely used titanium alloys due to its good overall performance, which accounts for more than half of the usage amount of all titanium alloys.10) Ti–6Al–4V alloy is a typical (α + β) two-phase titanium alloy and of its strength is approximately 900 MPa.11) According to the literature,3,4) the mechanical properties of some titanium alloys can be improved by adding alloying elements. Some researchers have proven that Fe is an effective eutectoid β-stabilizing and grain refinement element.12–14) As an important constituent in high-strength titanium alloys, Fe has been used in the early commercially significant β alloy Ti–10V–2Fe–3Al (Ti1023),15) Ti–4.5Fe–6.8Mo–1.5A1 (Ti-LCB),16) to Ti–6V–6Mo–6Fe–3Al (Ti6663).17) In addition, Fe is very cheap. We have applied a combinatorial approach to investigate the influence of Fe content on the microstructures and properties of Ti–6Al–4V alloy. We found that Ti–6Al–4V–5Fe alloy formed a very fine structure, which indicated a pseudo-spinodal decomposition has happened. The work of Andrew Boyne et al. has also described the Pseudospinodal mechanism of fine alpha/beta microstructure in β-Ti alloys.18) A relatively inexpensive near-β titanium alloy with high tensile strength and moderate elongation after (α + β) solution and aging treatments, Ti–6Al–4V–5Fe alloy has been designed recently. In general, titanium alloys are often subjected to heat treatments to achieve improved mechanical properties. The first thing to do in the heat treatment processes is the solution treatment, and achieving the desired microstructure and mechanical properties is extremely important.19–21) Ti–6Al–4V–5Fe alloy, like other near-β alloys, is thermomechanically deformed above the β transus temperature and then finishes forged below the β transus temperature. The alloy is treated above or below the β transus temperature to change the morphology and volume fraction of αp phase and then aged at different low temperatures to obtain precipitates of fine αs phase. The morphology, volume fraction, size and aspect ratio of αp are very important factors for the mechanical properties of the titanium alloy.22–26) The relation between the microstructure and the property of Ti–6Al–4V–5Fe alloy has been studied systematically, and the optimum solution and aging temperatures are reported in this paper.
The Ti–6Al–4V–5Fe alloy was double melted by vacuum arc remelting. The ingot with 160 mm diameter was forged at around 1050°C to a rod, and the deformation was approximately 50%. The rod was then forged for the second time at around 900°C to a square rod with dimensions of 1000 × 80 × 40 mm. The square rod was then cut into blocks with dimensions of 80 × 40 × 10 mm and was forged for the third time at around 850°C to a square rod with dimensions of 320 × 10 × 10 mm. The approach curves for the present Ti–6Al–4V–5Fe alloy were determined through metallographic method by measuring the area fraction with the OLYMPUS/PMG3 optical microscope using a series of metallographic samples with dimensions of 10 × 10 × 10 mm. The β transus temperature of alloy is approximately 875°C. Heat treatments were carried out in a tube furnace. The samples of the as-forged alloy were treated in a preheating furnace at the two-phase region and the β single-phase region (700°C, 740°C, 780°C, 820°C and 900°C) and held for 0.5 h then air cooled (AC). The AC samples were then aged at 500°C, 550°C and 600°C for 6 h in the preheated furnace and subsequently AC. Specimens for the optical microscope were mechanically polished by standard metallographic methods. The microstructure analysis of the samples was performed on a JEOL-JSM 7001F Field Emission Gun-Scanning Electron Microscope, which was operated at 20 kV voltage and Tecnai G2 20 ST transmission electron microscopy (TEM), which was operated at 200 KV. Firstly, the TEM thin foils were grounded to 50 µm thick on waterproof abrasive papers and then prepared with MTP-1A twin jet electrochemical polishing method at −30°C in an electrolyte consisting of perchloric, n-butyl alcohol and methanol (volume ratio 6%:34%:60%). A strain-gauge was attached to the samples, which were subjected to tensile tests. The fracture surfaces of the failed samples were examined by a field-emission scanning electron microscopy (SEM) system.
The chemical composition of the Ti–6Al–4V–5Fe alloy was analysed and is listed in Table 1. Using the empirical equation, the molybdenum equivalent of the Ti–6Al–4V–5Fe alloy was found to be approximately 12.9.
The microstructure of the forged sample is presented in Fig. 1. From Fig. 1(a), a large number of dark spherical α phase can be observed. In addition, the broken β grain and some subgrain boundaries are also clearly seen in Fig. 1(b). From the X-ray diffraction, the microstructure of alloy forged was found to comprised α and β phases. When the titanium was forged with large deformation below the α/β transformation point, more spherical α phase will occur. The spherical α phase has strong influence on the mechanical properties of the alloy.
Microstructure and XRD pattern of the as-forged Ti–6Al–4V–5Fe alloy.
The microstructures of the sample solution treated above or below the β transus temperature at five different temperatures, namely, 700°C, 740°C, 780°C, 820°C and 900°C for 0.5 h are shown in Figs. 2(a), (b), (c), (d) and (e), respectively. The microstructures of the samples treated in the two-phase region and β single-phase region are not uniform. Solution treatment below the β transus temperature, i.e., between 700°C and 820°C, produced a β matrix, which distributes some αp particles. As seen in Fig. 2(a), when the sample was treated at 700°C, a large number of elongated αp-phases were distributed in the matrix, and only a few globular αp-phases were distributed in the original grain boundary. When the solution temperature increased to 740°C, the aspect ratio of elongated αp-phases decreased. Meanwhile, the number of globular αp-phase distributed on the original grain boundary increased [Fig. 2(b)]. The number of globular αp-phase after treatment at 780°C was more than that of the elongated αp-phase [Fig. 2(c)]. When the alloy was treated at 820°C, no elongated αp-phases were found; the number of the total αp-phase decreased, and the size of the globular αp-phase increased [Fig. 2(d)]. The results indicated that the morphologies of the αp-phase were significantly affected by the solution temperature. When the alloy was treated above the β transus temperature, no αp-phase was observed [Fig. 2(e)].
Microstructures of the Ti–6Al–4V–5Fe alloy after treatment at different temperatures for 30 min followed by air cooling: (a) 700°C, (b) 740°C, (c) 780°C, (d) 820°C, and (e) 900°C.
Figures 3(a) and (b) show that the α′′ plates can be formed when the sample was quenched at 900°C. The electron diffraction pattern of the selected region is shown in (c), and the schematic is shown in (d). When α′′ martensitic transformation occurs, the [100] β direction shrinks to form [100] α′′, and the [011] β direction expands to form [010] α′′.27,28) The orientation relationship between the β and α′′ phase determined in this study is equivalent to those proposed by other researchers.27,29,30) The lattice correspondence between β and α′′ phases can be described as follows: [100]α′′ // [100]β, [010]α′′ // [011]β, (002)α′′ // $(0\bar{1}1)_{\beta }$, $(00\bar{2})_{{\alpha ''}}$ // $(01\bar{1})_{\beta }$. In the sample, which was quenched at the two-phase region, a significantly coarse α phase was formed, and α′′ martensite cannot be recognised. When the alloy contained a large number of α′′ phase, its tensile plasticity was low.
Formation of α′′ martensite plates by quenching at 900°C. (a) Optical microstructures; (b) Bright field image showing large α′′ plates; (c) selected area electron diffraction pattern and (d) key diagram.
The microstructures of the Ti–6Al–4V–5Fe alloys that were treated at different temperatures (700°C, 740°C, 780°C, 820°C) for 0.5 h and aged at 500°C for 6 h are shown in Fig. 4. Figure 4(a) shows few globular αp-phases and a large number of elongated αp-phases distributed in the β matrix. Thus, nearly no αs phases were observed because the low solution temperature resulted in the higher stability of the β matrix and the lower driving force of the αs phase formation during aging. When the solution temperature was 740°C, the globular αp-phases increased, the elongated αp-phases decreased in Fig. 4(b), and few αs phases were seen. When the solution temperature was 780°C, a large number of globular αp-phases, some αs phases and few elongated αp-phases can be seen from Fig. 4(c). However, when the alloy was treated at 820°C, only some globular αp-phases, without elongated αp-phases, and more αs phases were found to be distributed in the β matrix [Fig. 4(d)]. The volume fractions of αp-phase, elongated αp-phase and globular αp-phase were strongly influenced by the solution temperature. Figure 4(e) shows that the volume fraction of total αp decreased from 60% after treatment at 700°C to 35% after treatment at 820°C. The proportion of the globular αp-phase to the αp-phase increased, and the total αp-phase decreased with the increase of solution temperature according to Fig. 4(e).
Microstructures of the Ti–6Al–4V–5Fe alloy after treatment at various temperatures plus the same aging condition (500°C/6 h/AC): (a) 700°C, (b) 740°C, (c) 780°C and (d) 820°C.
To determine the distributions of the αs phase, the samples were observed by TEM. The typical bright-field images and the corresponding results of SADP are illustrated in Fig. 5 and Fig. 6. The αs phase can be clearly seen in the TEM photograph [Figs. 5(a), (c), (e) and Fig. 6(a)]. The αs plates have mixed sizes at nanoscale of 100 nm and at microscale of 0.5 µm [Figs. 5(a), (c), (e)]. In the treated at 820°C and then aged at 550°C, acicular αs phases with sizes in the range of 0.5–1.5 µm were observed [Fig. 6(a)] to be distributed in β the matrix. Meanwhile, the αs plates demonstrated an acicular shape with two orientations intersecting at an angle of approximately 60°. Figure 5(b) shows a $[\bar{1}13]$ β zone axis electron diffraction pattern from the sample after treatment at 740°C and then aging at 500°C. The microstructure consisted of both β phase and small α particle, whose diffraction maxima approximated to 1/2 {112} β highlighted by the red ellipse in the SAD. In addition, Figs. 5(d), (f) and Fig. 6(b), correspond to the selected area electron diffraction patterns that recorded the [011] β, [001] β and $[\bar{1}11]$ β zone axes from different samples.
TEM micrographs and SADPs of Ti–6Al–4V–5Fe alloy after different heat treatments: (a) 740°C, 500°C (c) 780°C, 500°C and (e) 820°C, 500°C. (b), (d) and (f) are SADPs corresponding to the α + β areas (a), (c) and (e) respectively.
(a) TEM micrographs of Ti–6Al–4V–5Fe alloy after heat treatments: 820°C, 550°C; (b) SADP corresponding to (a).
Figure 7 shows the microstructures of the Ti–6Al–4V–5Fe alloy treated at four different temperatures (700°C, 740°C, 780°C and 820°C) and aged at 550°C for 6 h. The microstructure shown in Fig. 7 was similar to that in Fig. 4. As Fig. 7(a) illustrates, a small amount of spherical αp-phases was distributed at the grain boundary, and the β grain size was dozens of micron. In Fig. 7(b), the microstructure consisted of β matrix and a certain amount of αp-phases. The number of globular αp-phases in Fig. 7(b) was more than that in Fig. 7(a), and the size of elongated αp-phases in Fig. 7(b) was larger than that in Fig. 7(a). This phenomenon is pronounced for this experiment. When the solution temperature increased from 740°C to 780°C, more globular αp-phases were distributed in the matrix [Fig. 7(c)]. Figure 7(d) shows that the volume fraction of total αp-phases decreased compared with that in Fig. 7(c). Given the high solution temperature (820°C), the adjacent globular α-phases have grown together to form an elongated morphology [Fig. 7(d)]. The αs phases distributed in the β matrix can be easily seen from Fig. 7(c), (d). Similar with Fig. 4, the volume fractions of elongated αp-phase and globular αp-phase were also counted in Fig. 7(e). The total αp volume fraction dropped from 63% after solution treated at 700°C to 28% after solution treated at 820°C. The total αp-phase decreased, but a proportion of the globular αp-phase to the αp-phase increased with increasing solution temperature, when solution temperature below 800°C.
Microstructures of the Ti–6Al–4V–5Fe alloy after treatment at various temperatures plus the same aging condition (550°C/6 h/AC): (a) 700°C, (b) 740°C, (c) 780°C and (d) 820°C.
Figure 8 illustrates the results of the EPMA of the surface of typical heat-treated samples. The heat treatments determine the distribution of the alloying elements, Al, V and Fe.31) A few lamellar α phases were formed on the grain boundaries (Fig. 8(a), which was solution treated at 700°C and then aged at 500°C), which was rich in Al and lean in Fe and V [Figs. 8(b) (c) and (d), respectively]. In addition, some globular αp-phase and elongated αp-phase can be seen in Fig. 8(e) and Fig. 8(i), which was solution treated at 780°C and 820°C then aged at 500°C. Figures 8(f)–(l) demonstrate that more Al-rich, Fe-lean and V-lean domains (more red and dark blue regions) were observed on the grain boundaries. Fe, an effective eutectoid β-stabilizing element, can improve the intensity of alloys dramatically. Figure 8(i) clearly shows precipitate-free zone (PFZ). It is well known that there are two reasons for the formation of the PFZ regions. The first reason is the vacancy depletion, and the second reason is the solute depletion.32–36) Since the aging temperature in the experiment is high (500°C), the reason for the formation of PFZ is likely to solute depletion. The grain boundary is a preferred location for nucleation of the precipitate. The precipitate can first nucleate at these sites by drawing solutes from adjacent matrix. Therefore, due to insufficient solute content, the area close to the grain boundary remains free of precipitation.35) The PFZ is detrimental to the properties of the alloy in most cases because the yield strength of the PFZ is low, and under stress, the plastic deformation tends to concentrate in the PFZ, causing intergranular fracture. Therefore, the elongation of the alloy solution treated at 820°C and then aged at 500°C is the lowest (2.5%).
Results of the elemental mapping of the surface of alloy samples that were heated at different temperatures. (a) (e) and (i) was solution treated at 700°C, 780°C and 820°C, respectively, then aged at 500°C. (a) (e) and (i) SEM images. (b)–(d), (f)–(h) and (j)–(l) EPMA elemental maps of Fe, V and Al, respectively.
The tensile results of Ti–6Al–4V–5Fe alloys after different heat treatments are presented in Fig. 9. The tensile results showed that the mechanical properties of the alloy were apparently dependent on aging temperature and solution temperature. The alloy, which was solution treated at 820°C and then aged at 500°C, demonstrated the highest ultimate strength of ∼1500 MPa, and the elongation of the alloy solution treated at 780°C and then aged at 600°C is the highest (13%). From the true stress–strain curves [Fig. 9(a)], the tensile strength of Ti–6Al–4V–5Fe alloys, which were solution treated at different temperatures (700°C, 780°C and 820°C) and then aged at 500°C increased from 1250 MPa to 1550 MPa, but the elongation of the Ti–6Al–4V–5Fe alloys decreased from 8.5% to 2.5%.
Stress–strain curves and tensile properties of Ti–6Al–4V–5Fe alloys after different heat treatments ((a) stress–strain curves; (b) aged at 500°C, (c) aged at 550°C and (d) aged at 600°C).
The different solution temperatures are illustrated by a set of typical aging curves [Figs. 9(b), (c), (d)]. As seen from Figs. 9(b), (c), (d), the trends of tensile strength of Ti–6Al–4V–5Fe alloy are similar. As the temperature of the solution increases, the tensile strength first decrease and then increase. In Figs. 9(b), (c), (d), the relation between elongation and tensile strength is inverse at the same aging temperature. The trends of elongation of Ti–6Al–4V–5Fe alloy that was subjected to different heat treatments were also similar. The alloy solution treated at low temperature (700°C, 740°C and 780°C) and then aged at different temperatures was hardened to 1300 MPa of the ultimate strength with 7%–13% elongation. The alloy solution treated at high temperature (820°C) and then aged at different temperatures was hardened to 1300–1500 MPa of the ultimate strength, but the elongation of the alloy was low. The mechanical properties of Ti–6Al–4V–5Fe alloy solution treated below 800°C and then aged were better than that of the solution treated above 800°C and then aged. The alloy solution treated at 780°C and then aged at 500°C exhibited high ultimate strength of up to 1362 MPa with 8.2% elongation.
The tensile properties of the titanium alloys can be strongly affected by this type of microstructure. The morphology of precipitated phase, such as size and shape, depends on aging and solution temperature. Therefore, the tensile properties of the alloy are also strongly influenced by heat treatment. Different solution and aging treatments lead to different mechanical properties.
As seen in Fig. 9(b), when solution treated at different temperatures (700°C, 740°C and 780°C) and then aged at 500°C, an increase in the elongation with the increase of solution temperature was observed. The elongation of the alloy was significantly greater than that of the alloy solution treated at 820°C and then aged at 500°C. As seen in Fig. 4, the number of globular αp-phases increased with the increase of solution temperature (700°C, 740°C and 780°C). When the solution was treated at 820°C, though the number of globular αp-phases did not reduce, the volume fraction of total α decreased. A decrease in driving force of the nucleation and growth of α phase was observed because the treatment temperature was close to the β transus temperature. To some extent, the ductility of the alloy can increase with the content of αp phase. That is, the αp-phase can improve the ductility to some extent.37) In addition, the effect of the globular αp-phases on ductility is better than that of the elongated αp-phases.13) The reason is that the elongated αp-phases, relative to the globular αp-phases, can reduce the planar spacing between particles, which then lead to higher hole density during deformation.13) Therefore, the elongation of Ti–6Al–4V–5Fe alloy first increase and then decrease with increasing solution temperature.
The strength and elongation trends are roughly opposite. The strength of β alloys can be improved with a reduction in ductility by heat treatments, which led to precipitations of the α particle. The fine α precipitates distributed in the β matrix can influence the movement of dislocations. The morphology of α particles strongly affect the strength and ductility of the alloy. In Fig. 9(b), the strength first decrease and then increase with increasing solution temperature. The effect of the globular αp-phases on strength is less than that of the elongated αp-phases.13) Obviously, the proportion of the elongated αp-phases to the αp-phases in Fig. 4(a) is larger than that in Fig. 4(b), but the proportion of globular αp-phases in Fig. 4(a) is smaller than that in Fig. 4(b). Therefore, the ultimate strength of the alloy solution treated at 700°C is higher than that of the alloy solution treated at 740°C. Obviously, with the increase of solution temperature, the solution of the alloying elements can be increased, which facilitates the dispersive precipitation of fine αs during aging.38) When the solution temperature was lowered, the driving force for the nucleation and growth of the fine αs phase during aging reduced. The critical precipitate nuclei size is inversely proportional to the driving force for the nucleation and growth of the precipitation.39) As shown in Fig. 5(a), (c), the lenticular αs in Fig. 5(c) is considerably more dispersed than that in Fig. 5(a). Thus, the ultimate strength of the alloy solution treated at 780°C is significantly greater than that of the alloy solution treated at 740°C. Similarly, the strength of the alloy solution treated at 820°C is higher than that of the alloy solution treated at 780°C according to Fig. 5(e). Therefore, the strength of Ti–6Al–4V–5Fe alloy first decreased and then increased with increasing solution temperature.
The αs phase in Fig. 6(a) is considerably coarser compared with that in Fig. 5(e). The primary reason for that may be two aspects. Firstly, the lower aging temperature reduces the driving force for the nucleation and growth of the fine αs phase during aging. The higher aging temperature in Fig. 6(a) offered sufficient driving force for the α phase. Secondly, the αp-phase restrains the growth of the fine β grains, which are beneficial to the limitation of αs phase.40) According to Figs. 4(d) and 7(d), the volume fraction (35%) of αp phase solution treated at 820°C and then aged at 500°C is more than that of the (28%) solution treated at 820°C then aged at 550°C. Thus, the size of β grain Fig. 4(d) is smaller than that shown in Fig. 7(d). In Fig. 6(a), the αs plates exhibit an acicular shape with two orientations intersecting at an angle of approximately 60°. The fine and crossed α plates have considerably enhanced effect because of their different sizes and orientations, which lead to high strength levels.40)
Figure 9(c) presents the trend of the elongation and strength of the alloy solution treated at different temperatures (700°C, 740°C, 780°C and 820°C) and then aged at 550°C, which is similar to that in Fig. 9(b). In Figs. 7(a), (b), (c) and (d), the globular αp-phases and elongated αp-phases are distributed at the grain boundary and grain interior. The proportion of the globular αp-phases to the αp-phase in Fig. 7(c) is the highest and that of the elongated αp-phases to the αp-phase in Fig. 7(a) is the highest according to Fig. 7(e). The β grain size is dozens of micron range because a large number of the αp limited β grain growth.
The fracture surfaces of some solution-treated and aged samples were studied, exposing a potential fracture mechanism for controlling the tensile properties. The plasticity of Ti–6Al–4V–5Fe alloy can be improved by optimising the microstructure, i.e., the morphology and amount of αp-phase, by means of suitable heat treatment.
Figures 10(a) and (b) show that the alloy fracture, which was solution treated at 820°C for 0.5 h and aged at 550°C for 6 h, mainly exhibits intergranular facture, except for minor ductile fracture. Figure 10(b) is a high magnification fractograph that clearly shows some shallow and fine dimples surrounding the facets. A possible explanation is that the main factor leading to low ductility is the α film located in the grain boundary [Figs. 7(d)]. The α film, located in the grain boundary, resulted in the deformation of some areas preferentially during plastic deformation, thereby separating the grains.16,17) Figures 10(c), (d) show the fracture surface of alloy, which was solution treated at 700°C for 0.5 h and aged at 550°C for 6 h. When the solution temperature decreased to 700°C, the fracture still appeared to be of mixed mode type, but the area fraction of faceted regions was lower, and the dimpled regions appeared to increase. Figure 10(d) is a high magnification fractograph that clearly demonstrates deep and fine dimples. Fracture surface of alloy solution treated at 780°C for 0.5 h and aged at 550°C for 6 h demonstrated a full ductile mode of tensile fracture [Figs. 10(e) and (f)]. The high-magnification fractographs of Fig. 10(f) show fine and deep dimples. The main deformation mode for the two-phase region solution treated and then aged was a strain localization in the β matrix, resulting in voids at the interface between β matrix and primary α-phase.38) The deep and fine dimples corresponded to α layer at the grain boundary.
SEM images of fracture surfaces of the alloy after solution treatment at various temperatures and the same aging condition (550°C): (a), (b) 820°C, (c), (d) 700°C and (e), (f) 780°C.
In this experiment, the microstructures and mechanical properties of Ti–6Al–4V–5Fe alloy strongly depend on heat treatment. The alloy was heated with different solution conditions (700°C, 740°C, 780°C, 820°C and 900°C) and then aged at temperatures ranging from 500°C to 600°C afterwards. The important results are summarised as follows:
The work was financially supported by National Key Technologies R & D Program of China (Grant no. 2016YFB0701301), National Natural Science Foundation of China (Grant No. 51671218, 51501229), National Key Basic Research Program of China (973 Program) (Grant no. 2014CB644000), and State Key Laboratory of Powder Metallurgy, Central South University, Changsha, China.
