Effects of β treatments on microstructures and mechanical properties of TC4-DT titanium alloy

Effects of β treatments on microstructures and mechanical properties of TC4-DT titanium alloy

Materials Science and Engineering A 533 (2012) 55–63 Contents lists available at SciVerse ScienceDirect Materials Science and Engineering A journal ...

4MB Sizes 1 Downloads 65 Views

Materials Science and Engineering A 533 (2012) 55–63

Contents lists available at SciVerse ScienceDirect

Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea

Effects of ␤ treatments on microstructures and mechanical properties of TC4-DT titanium alloy Xiaona Peng ∗ , Hongzhen Guo, Tao Wang, Zekun Yao School of Materials Science and Engineering; Northwestern Polytechnical University, Xi’an 710012, PR China

a r t i c l e

i n f o

Article history: Received 1 September 2011 Received in revised form 12 November 2011 Accepted 12 November 2011 Available online 22 November 2011 Keywords: TC4-DT titanium alloy Near-isothermal forging Microstructure Tensile properties Fracture toughness

a b s t r a c t ␤ Processing (deformation in ␤ phase field followed by heat treatment in ␣ + ␤ phase field) and ␤ annealing (deformation in ␣ + ␤ phase field followed by annealing in ␤ phase field) were carried out to research their influence on microstructures and mechanical properties including fracture toughness of TC4-DT titanium alloy. The tensile properties at room and high temperature as well as fracture toughness were tested for all the experiment conditions. The microstructure evolution and fracture surfaces were researched by optical microscope and scanning electronic microscope (SEM) and the microstructure features were measured by means of image analysis software. Results showed that the microstructures were lamellar in ␤ processing and acicular Widmanstatten in ␤ annealing respectively. Spheroidization of ␣ lamellar was found in the microstructures of ␤ processing. SEM observation showed that the fracture mechanism changed from transcrystalline in the ␤ processing conditions to a mixture of intercrystalline and transcrystalline at the ␤ annealing conditions. The tensile strength and plasticity did not change much under the ␤ processing conditions. While at ␤ annealing conditions, the strength and plasticity varied with the temperature in a reverse trend. The biggest fracture toughness was obtained at ␤ annealing conditions. It was found that ␤ annealing was preferable to ␤ processing with regard to obtaining high fracture toughness and tensile properties with a little sacrifice of plasticity which does not affect its practice use. © 2011 Elsevier B.V. All rights reserved.

1. Introduction TC4 is an ␣ + ␤ titanium alloy widely used in the aerospace industry and biomedical due to its highly attractive properties such as good formability, low density, high specific strength, excellent corrosion resistance and high temperature strength retention [1,2]. TC4-DT is another titanium alloy produced based on TC4 with the volume fraction of C, N, O in a controlled level at an expectation to attain high fracture toughness. Its properties are equivalent to those of Ti–6Al–4VELI. The mechanical properties of alloys depend strongly on the chemical composition, processing history and heat treatment procedures which decide the varieties of microstructures [3,4]. There have been two other ways to obtain high fracture toughness besides the control of chemical composition, viz. ␤ processing with the deformation temperature above the ␤ transus and ␤ annealing with the annealing temperature above the ␤ transus reported in recent literature [5]. However there has no systematic research on both these two ways and the comparison between them. The aim of this work is

∗ Corresponding author. Fax: +86 29 88493744. E-mail address: [email protected] (X. Peng). 0921-5093/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2011.11.033

to investigate the influence of ␤ processing and ␤ annealing on the microstructures and mechanical properties especially fracture toughness of TC4-DT, with a view to get a preferable processing for the balance of fracture toughness, strength and ductility. Previous studies [3,6,7] have shown that processing of ␣ + ␤ titanium alloys above ␤ transus results in lamellar microstructure morphology, consisting of ␣ platelets with an inter-platelet ␤ phase. Thus the microstructure evolution during deformation and heat treatment in ␤ phase field is also investigated with the aid of optical microscope and SEM. The microstructure parameters including the volume fraction of the ␣- and ␤ phase, the primal ␤ phase size, the thickness of lamellar ␣ and the characteristics of grain boundary ␣ are quantified by means of the image analysis software.

2. Materials and experimental procedure 2.1. Materials TC4-DT used in this research was received in hot-rolled bar with a diameter of ϕ 215 mm from which cuboid specimens (30 mm × 14 mm × 66 mm for tensile test and 66 mm × 40 mm × 40 mm for fracture toughness test) were

56

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

2.2. Experimental procedures

Fig. 1. Microstructure of TC4-DT alloy hot-rolled bar.

spark machined. The microstructure of the hot-rolled bar is shown in Fig. 1. It is a fully homogeneous equiaxed structure consisting of 70% primary ␣ phase with the average grain size 9 ␮m and transformed ␤ with the secondary lamellar ␣ thickness 1.4 ␮m.

The ␤ transus temperature (T␤ ) of the material obtained by examination of microstructure is 975 ◦ C. The near isothermal ␤ processing was carried out on a computer-controlled 630 ton hydraulic press which allowed the specimens to be pressed at constant strain rates. The temperature for ␤ processing was 995 ◦ C and 1010 ◦ C with the same deformation extent 60% and constant strain rate 1 × 10−3 s−1 , followed by air cooling (AC). The heat treatment was 800 ◦ C/1 h, AC. At ␤ annealing experiments, the primary material were isothermal forged at 945 ◦ C at the condition of 60% deformation degree and constant strain rate 1 × 10−3 . The specimens were then heated to the first annealing temperature 985 ◦ C, 995 ◦ C, 1005 ◦ C for 40 min and air cooled. The aging treatment was 730 ◦ C/2 h, AC. The experimental scheme is given in Table 1. During the deformation process, the specimens were heated to the designated temperatures and soaked for 12 min for tensile test and 32 min for fracture toughness test respectively before forging. A borosilicate glass lubricant was spread over on the surface of the specimens. The forging dies were heated by cylinder-shaped electric resistance furnaces fixed in the hydraulic press. The tensile tests were executed on the ENST-1196 stretcher at room temperature and 400 ◦ C separately. Cylindrical specimens with a gauge length of 71 mm and a diameter of 5 mm were employed. The results were the mean values obtained from

Fig. 2. High (left) and low (right) magnification optical microstructures of TC4-DT alloy after tensile test at room temperature (a) 995 ◦ C deformed, AC followed by 800 ◦ C/1 h, AC; (b) 1010 ◦ C deformed, AC followed by 800 ◦ C/1 h, AC.

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

57

Table 1 Experimental scheme. Process no.

Temperature of dies

Deformation condition

Heat treatment

1 2 3 4 5

950 ◦ C

995 ◦ C, 60% 1 × 10−3 s−1 AC 1010 ◦ C, 60% 1 × 10−3 s−1 AC 945 ◦ C, 60% 1 × 10−3 s−1 AC 945 ◦ C, 60% 1 × 10−3 s−1 AC 945 ◦ C, 60% 1 × 10−3 s−1 AC

800 ◦ C/1 h, AC 800 ◦ C/1 h, AC 985 ◦ C/30 min, AC + 730 ◦ C/2 h, AC 995 ◦ C/30 min, AC + 730 ◦ C/2 h, AC 1005 ◦ C/30 min, AC + 730 ◦ C/2 h, AC

945 ◦ C

two tested specimens. The fracture toughness tests using compact tension were carried out on an INSTON1252 servohydraulic testing system in accordance with ASTM E399 standard for damage tolerance testing. The microstructure was observed using the OLYMPUSPM-G3 microscope. The samples were etched with 3% HF + 6% HNO3 + 91% H2 O solution. The fracture surfaces of specific specimens were observed using SEM SUPRA55.

3. Microstructure evolution 3.1. Microstructures under ˇ processing Representative material microstructures taken from the ␤ processing specimens at different temperatures are shown in Fig. 2. It can be seen that both of the microstructures consist of primary ␣ lamellar and transformed ␤ (␤t ) that consists of acicular ␣ delineated by ␤. With the temperature increasing, the content of primary

␣ lamellar increases and the morphology also shows difference. From the low magnification micrographs in Fig. 2 (upper and lower right), the primary ␤ grains are elongated along the flow direction. The microstructure evolution under ␤ processing of the processing no. 1 is shown in Fig. 3. It consists of three stages: deformation in the ␤ phase field, cooling to room temperature from ␤ phase field and 800 ◦ C/1 h, AC heat treatment. The microstructure of the specimen deformed at 995 ◦ C was retained by water cooling after deformation. For comparison, one additional sample was heated to 990 ◦ C, and then quenched in water without deformation (Fig. 3a). As seen in Fig. 3a when the temperature was above the ␤ transus (990 ◦ C) all the primary equiaxed ␣ were transformed into high temperature ␤ phase before hot working, so only ␤ phase was subjected to deformation during the forging in ␤ phase field. The ␣ phase at prior ␤ grain boundary is not distinct on postdeformation quenching or air cooling because of the broken up induced by large deformation degree, meanwhile the distorted and straight lamellar ␣ also appear out, as shown in Fig. 3b and c.

Fig. 3. The microstructures in different stages of ␤ processing at 995 ◦ C (a) the microstructure at 995 ◦ C without deformation; (b) deformed in ␤ phase field followed by water cooling; (c) deformed in ␤ phase field followed by air cooling (AC); (d) 800 ◦ C/1 h, air cooling heat treatment.

58

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

Table 2 Average thickness of the prior ␣, secondary ␣ lamellar and the volume fraction of phases at different ␤ processing conditions unit: ␮m. Process no.

Distribution of aspect ratio of prior ␣ lamellar

Primary ␣ thickness

Secondary ␣ thickness

Volume fraction of ␤t (%)

1 2

Dispersive Concentrated

3.00 3.79

0.7877 0.6466

36 12

The mechanism of microstructure evolution in ␤ phase is involved dynamic recovery and dynamic (or meta dynamic) recrystallization [3] which can be validated by the flow softening of Ti–6Al–4V–ELI deformed in the ␤ phase field [4]. There are three possible mechanisms responsible for that phenomenon [4,8–10]: (a) adiabatic heating during hot pressing; (b) ␣ → ␤ phase transformation during hot pressing and (c) dynamic recovery and recrystallization (DRX) in different deformation region, whereas the dynamic recovery cannot be confirmed by microstructure observation. While Liu and Baker [11] found that dynamic recovery is dominant during the forging of the alloy IMI685 in the ␤ field which is the case in this experiment. From the microstructure shown in Fig. 3b and c, it can be seen that the recrystallized prior ␤ grain is not clearly found out suggesting that DRX is not intense. It can also figure out that the cooling rate only influences the width of the lamellar ␣ or Widmanstatten. The microstructure after heat treatment of 800 ◦ C/1 h, AC is shown in Fig. 3d. It consists of primary ␣ lamellar and transformed ␤ structure, and the width of the secondary ␣ in ␤t was smaller than that of the primary ␣ lamellar. This can be explained that when the deformed materials are heated to 800 ◦ C and soaked for 1 h, the ␣ → ␤ phase transformation will take place and this is more easily for the thinner and distorted lamellar ␣ resulted by the large stored energy. After that these ␤ phases will transform into secondary ␣ lamellar during the subsequent air cooling. The spheroidization of the prior ␣ lamellar is displayed in Fig. 3c and d. According to Weiss et al. [12], the possible mechanisms for the separation of the lamellar ␣ phase to shorter segments are the formation of sub-boundaries across ␣ plate, the localized shear and rotation of the lamellar during the hot deformation. The driving force for spheroidization is the surface energy reduction [3] which enhances the ␤ phase penetrate into ␣ phase and partially or fully separates the ␣ phase into shorter segments. In this experiment we can infer that the deformation in ␤ phase field induced much inner-stress such as shear stress and dislocation (dislocation cell), during the following air cooling and under the action of this shear stress, the dislocation gliding and atomic diffusion get started which can result in spheroidization, and such spheroidization of the prior ␣ lamellar is shown in Fig. 4a. While in the heat treatment, the dislocation cell will become the nucleation site and the extended soaking time at high temperature supply enough motivation for the atom diffusion, thus the recrystallization takes place as shown in Fig. 4b. The different mechanisms are shown in Fig. 4. The mechanism of the microstructure evolution during the deformation at 1010 ◦ C is similar to that forming at 995 ◦ C. The differences lie on the width of the primary ␣ lamellar and second ␣ lamellar, the distribution of the aspect ratio of the prior ␣, the width of ␤ grains and the volume fraction of primary ␣ lamellar and transformed ␤ structure (␤t ). Table 2 shows the values of these parameters. From Table 2 it can be seen that the average thickness of primary ␣ increased while the average aspect ratio decreased with increasing temperature. The thickness of the prior ␣ is influenced by many factors, such as the strain rate, local grain orientation, temperature, holding time [3] and cooling rate after deformation. Besides that, the evolution of the microstructure during the deformation in ␤ phase field is different with that without deformation because

of the dynamic equilibrium induced by the dynamic recovery or recrystallization and the strain hardening, and this also influences the thickness of the primary of ␣. The average thickness of the newly formed secondary ␣ in ␤t does not differ much at all temperatures compared to the primary ␣ thickness. While the volume fraction of ␤t decreases with the temperature increasing. This validates that at 800 ◦ C the content of the ␣ → ␤ phase transformation of the specimen forged at 1010 ◦ C was not as much as that of the one forged at 995 ◦ C. It can be explained that higher deformation temperature supplies higher energy for the occurrence of the dynamic recovery or recrystallization (at this experiment dynamic recovery dominated) and the microstructure obtained is more homogeneous and that is the reason for the concentrated distribution of aspect ratio of prior lamellar ␣ in Fig. 2b. During the subsequent heat treatment, the driving force for phase transformation is not as intense as that deformed at lower temperature.

Fig. 4. Different mechanism for the spheroidization of the prior ␣ lamellar for the ␤ processing at 995 ◦ C.

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

59

Table 3 Microstructure parameters at different temperatures of ␤ annealing unit: ␮m. Process no.

D

The ␣ lamellar width

la

The second ␣ lamellar width

3 4 5

785.25 829.82 772.42

1.78 1.84 1.84

1.83 2.32 1.80

0.8143 0.8592 0.8453

3.2. Microstructures under ˇ annealing Fig. 5 shows the microstructures at different temperatures of ␤ annealing. As seen in Fig. 5, the microstructures obtained in the ␤ annealing are typical ␤ transformed Widmanstatten structure. Prior-␤ grain size (D), the ␣ lamellar width and the second ␣ lamellar width, the ␣ colony size and the thickness of ␣ layers at ␤ grain boundaries (la ) to characterize the morphology of the microstructures at different temperatures of ␤ annealing are measured and shown in Table 3. Notably in between the colony boundary and the primary ␣ layer there is a thin layer of ␤ [4] which can be verified by the transmission electron microscope [13]. From Table 3 it can be seen that the prior-␤ grain size increases firstly and then declines with the increasing temperature of the heat treatment at ␤ phase field, and the thickness of ␣ layers at ␤ grain boundaries has the same trend. This is attributed to ␤ grain nucleation and growth kinetics in ␤ phase field which will be further studied in the future research, while other parameters do not differ much. The reason is that during the ␤ annealing the most important parameter is the cooling rate from the ␤ phase field [13] which determines the size of the ␣ lamellar, the ␣ colony size and the thickness of ␣ layers at ␤ grain boundaries. The second ␣ lamellar is mostly affected by the temperature of the heat treatment in ␣ + ␤ phase field and the following cooling rate. It should be noted that ␣ colony size is also influenced by the sides and size of prior ␤ polygon grains as discovered in this research. However due to the limited resolution ratio of the light microscope, the ␣ colony size is hard to determined. The microstructure evolution in the ␤ annealing is shown in Fig. 6. It can be seen from Fig. 6a that the microstructure deformed in ␣ + ␤ phase field is different from that in ␤ phase field. It consists of equiaxed ␣, elongated ␣ and transformed ␤ structure. When heated to ␤ phase field, recrystallization of ␤ phase occurred and ␤ grain size can be controlled by the selection of temperature [13]. Once air cooling to room temperature, the microstructure shows large ␤ grains with continuous ␣ layers at ␤ grain boundaries and ␣ colonies with different orientations in ␤ grains, as shown in Fig. 6b. The microstructure in the second heat treatment is similar to that of the first. The difference lies in the finer acicular ␣ emerging between the thick lamellar ␣ plates. 4. Mechanical properties 4.1. Tensile properties ◦

Fig. 5. Microstructures at different temperatures of ␤ annealing (a) 985 C; (b) 995 ◦ C; (c) 1005 ◦ C.

In alpha–beta titanium alloys the mechanical properties depend on a number of factors. These factors include phase chemistry and stability, volume fraction of the phases, strength ratio of the phases,

Table 4 Mechanical properties of TC4-DT alloy at different ␤ processing conditions. Process no.

Test temperature (◦ C)

UTS (MPa)

YS (MPa)

Elongation (%)

Reduction area (%)

1

Room temp 400

920 607.5

872.5 500

13.25 17.75

39.25 57.5

2

Room temp 400

920 612.5

862.5 510

13.25 15.25

41 57.5

60

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

Fig. 6. Microstructures at different stages of ␤ annealing (a)microstructure deformed at 945 ◦ C followed by air cooling; (b) microstructure of the first heat treatment at 995 ◦ C followed by air cooling; (c) microstructure of the second heat treatment at 720 ◦ C followed by air cooling.

Fig. 7. The tensile fractographs of TC4-DT alloy tested at different temperature (a) process no. 1 tested at room temperature; (b) process no. 5 tested at room temperature; and (c) process no. 5 tested at 400 ◦ C.

and the morphology of the phases, including the grain size and orientation relationship [14–16]. A summary of tensile tests at room temperature and 400 ◦ C temperature are presented in Tables 4 and 5. As shown in Table 4 ultimate tensile strength at room temperature of the two samples is constant at 920 MPa while the yield strength is 872.5 MPa and 862.5 MPa respectively. The higher yield strength of 872.5 MPa is

contributed to the more finer prior ␣ lamellar. According to [13], the yield strength is partially determined by the slip length (␣ colony size concerning the lamellar microstructures), and the decreased ␣ colony size will result in the reduction of slip length and a corresponding increase in yield strength. The tensile ductility at both conditions varies not much with the same elongation 13.25% and reduction in area 39.25% and 41% respectively.

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

61

Table 5 Mechanical properties of TC4-DT alloy at different ␤ annealing temperatures. Process no.

Test temperature (◦ C)

UTS (MPa)

YS (MPa)

Elongation (%)

Reduction area (%)

3

Room temp 400 Room temp 400 Room temp 400

927.5 610 912.5 597.5 932.5 597.5

840 480 827.5 475 855 477.5

7.5 15 9.0 12.75 6.25 19

11.5 43.25 19 45 19 65

4 5

While at high testing temperature 400 ◦ C, the strength and the tensile ductility are very different from the properties tested at room temperature. The strength declined drastically to more or less 610 MPa, but the tensile ductility was higher than that of the room temperature. The deformation mechanism is different from each other at room temperature and 400 ◦ C [17]. The critical stress to start dislocation decrease with increase in temperature, so the 400 ◦ C tensile strengths were lower than the respective ones tested at room temperature. It is well known that at high temperature, the gliding of grain or phase boundary dominates in the deformation mechanism [17,18] and that the distorted energy generated in the grain boundary is so large that at proper temperature and strain rate it will become softer as a result of recovery which relax the stress concentration and make the deformation continuously. So the tensile plasticity increased with the increasing test temperature for each specimen. Anyway, whether at room temperature or at 400 ◦ C, the variation of tensile properties are not distinct at different ␤ processing temperatures. The tensile properties of the samples after ␤ annealing at different temperature are listed in Table 5. As seen in Table 5, when tested at room temperature, the strength shows a trend of first declining and then rising. The ductility especially the elongation illustrates the opposite trend. While at 400 ◦ C the tensile strength declines with the increasing processing temperature, but the ductility increases with a fluctuation at the elongation of process no. 4. The highest ductility is 19% of elongation and 65% of reduction in area at the process no. 5. The highest tensile strength and high temperature ductility and lowest room temperature ductility at the process no. 5 may be attributed to smaller prior ␤ grain size and more grain boundary ␣. This indicates that the adverse influence of grain boundary ␣ exceed the positive influence of finer prior ␤ grain size on ductility. When comparing the tensile properties, the ultimate strength of ␤ annealing exceeds those of the ␤ processing with the exception 912.5 MPa at the process no. 4. To that contrary are the yield strength and ductility all of which are lower than those of ␤ processing whether at room temperature or 400 ◦ C testing. The reasons for the difference are complicated. According to [19] the tensile ductility is mainly determined by two parameters, i.e. crack nucleation resistance being the dominating parameter and crack propagation resistance. The former is strongly dependent on effective slip length (the ␣ colony size) [13], with the slip length decrease, the ductility reaches a maximum. Fracture modes change from transcrystalline dimple type of fracture to intercrystalline dimple type along the continuous ␣ layers at ␤ grain boundaries. These different fracture mechanisms between the ␤ processing and ␤ annealing at the room temperature test can be verified by SEM micrographs of fracture surfaces as shown in Fig. 7a and b (process no. 2 and no. 5 with the highest and lowest room temperature tensile ductility respectively). It can be seen that there are dimples in both Fig. 7a and b, but the dimples in Fig. 7b are very less than those in Fig. 7a, besides that there are tear ridges in Fig. 7b. From the observation of Fig. 7b it can be found that both of intercrystalline and transcrystalline fracture mechanisms exist. Gungor et al. [6] also pointed out that

colony boundaries and ␣ at prior ␤ grain boundaries may serve as void nucleation sites. Thus reducing the prior ␤ grain size can lead to an increase in tensile ductility. However in this experiment this is not the case because larger prior ␤ grain size reduces the grain boundary ␣ and this can lessen the void nucleation sites. The experiment results in this experiment suggest that the influence of ␤ grain boundary on ductility is more significant than colony boundaries. While the fracture surface of high temperature test specimen in ␤ annealing (process no. 5 with the highest high temperature ductility) is shown in Fig. 7c. The SEM micrographs of fracture surface exhibit typical ductile dimple fracture pattern due to less crystal defects at 400 ◦ C tensile testing. 4.2. Fracture toughness analysis The fracture toughness under different processing are listed in Table 6. As illustrated, all the fracture toughness is above 90 MPa m1/2 , and the fracture toughness of ␤ annealing are slightly higher than that of ␤ processing. It should be noted that among the different ␤ treatments the variation in their influence on the fracture toughness is not very much because of the lamellar microstructures. This also proves that the lamellar microstructure is more significant from the point of the fracture toughness [20]. It is well known that Widmanstatten ␣ structures exhibits greater toughness comparing with equiaxed ␣ structure, and coarsening of Widmanstatten ␣ structure is very effective to achieve higher toughness in ␣ + ␤ titanium alloys [21] as testified by the experiment process no. 4 in this research. Besides the width of Widmanstatten ␣, the width of grain boundary ␣, ␣ colony size or intersubcolony spacing also affect the fracture toughness. The relationship between fracture toughness and ␤ grain size is difficult to confirm according to Refs. [20,21] which have the contrary conclusion. The influence of ␤ grain size on fracture toughness can be ascribed to its effect to other microstructure parameters. The qualitative relationship between the prior ␤ grain size and other microstructure parameters is analyzed by Niinomi and Kobayashi [21] and the crack propagation path with the increasing of ␤ grain size is also illustrated in Fig. 8 [21]. It is shown that the number of intersubcolonies increases and the size of intersubcolonies

Fig. 8. Crack propagation path with the increasing of ␤ grain size [21].

62

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

Table 6 Fracture toughness under ␤ processing and ␤ annealing conditions. Process no.

1

2

3

4

5

KIC (MPa m1/2 )

91.0

93

94.8

96.9

95.8

decreases with increasing prior ␤ grain size. The crack deflection at subcolony boundaries shows a zigzag path which consumes more energy than straight path, leading to higher fracture toughness. This condition is appropriate when the crack follows an intercolony path [20]. The thickness of lamellar ␣ also poses a great effect on mechanical properties of the alloys. The optimal thickness of ␣ platelets is about 2–5 ␮m which leads to an optimum crack propagation resistance and toughness of the alloy [3]. The thickening of the alpha plates can increase fracture toughness by contributing to crack deviation. Such improvement by alpha plate coarsening has been observed in many previous investigations [21]. In the ␤ annealing conditions, the thickest ␣ lamellar corresponds to the largest fracture toughness in process no. 4, but this is not proper to the ␤ processing all of which the width of lamellar ␣ is thicker than that of ␤ annealing. In addition to the lamellar width, the fracture toughness also can be improved with increasing the amount of the lamellar ␣ phase [7]. In terms of grain boundary ␣, the increase of the thickness of ␣ at ␤ grain boundary has a positive effect on fracture toughness. It has been shown that structures containing thick grain boundary alpha layers consume more energy for crack propagation [22]. This can account for why the fracture toughness of ␤ annealing is higher than that of ␤ processing and the highest fracture toughness values corresponds to the thickest grain boundary alpha in process no. 4. Fig. 9 shows the SEM micrographs of fracture surfaces of the specimens in process no. 1 and process no. 4 in the fracture toughness test. It can be found that the fracture surface changes from the transcrystalline in Fig. 9a to both transcrystalline and intercrystalline fracture as shown in Fig. 9b. This can be attributed to the presence of the grain boundary ␣ layer in the ␤ annealing conditions. The high magnification in Fig. 9c depicts that the crack deflection and branching in ␤ annealing condition is more severe than that in Fig. 9a which means that there needs more work to be done in creating new surfaces and enables a larger volume of material to contribute to stress energy relief by plastic deformation [20]. This further demonstrates the high fracture toughness of the specimen in process no. 4. 5. Conclusion ␤ Processing and ␤ annealing were adopted to study their effect on the microstructures and mechanical properties of Ti–6Al–4V–ELI alloy. Optical microscopy and SEM were utilized to characterize the microstructures. Tensile test at room temperature and 400 ◦ C and fracture toughness test were conducted on the specimens at each condition to obtain the mechanical properties. According to the experiment results, the following conclusions can be attained.

Fig. 9. SEM fractographs of different conditions (a) transcrystalline fracture in process no. 1; (b) low magnification depicting both transcrystalline and intercrystalline fracture in process no. 4; (c) high magnification depicting crack deflection and branching in process NO.4.

(1) At the ␤ processing conditions, the microstructures were mixture of the primary lamellar ␣ and transformed ␤ microstructures. The grain boundary was not distinct as a result of broken up during the deformation. The spheroidization of the prior ␣ lamellar occurred due to different mechanisms at all the ␤ processing conditions. Microstructure evolution was not different at the two temperatures, the difference lay in the

X. Peng et al. / Materials Science and Engineering A 533 (2012) 55–63

different parameters of the microstructures. While at the ␤ annealing conditions, the microstructure was typical ␤ transformed Widmanstatten structure consisting of ␣ at prior ␤ grain boundaries and transformed ␤ interiorly. (2) Whether at room temperature or at 400 ◦ C, the variation of tensile properties and ductility were not distinct at different ␤ processing temperatures. Both of the strength was 920 MPa with the same elongation 13.25%. The reduction in area was 39.25% and 41% respectively. The strength at 400 ◦ C declined drastically to 610 MPa more or less, while the tensile ductility and particularly reduction in area were higher than those of the room temperature. At the ␤ annealing conditions, the strength tested at the room temperature showed a trend of first declining and then rising. The ductility especially the elongation illustrated the opposite trend with the highest of elongation and reduction in area 9.0% and 19% respectively. While the tensile strength at 400 ◦ C declined with increasing ␤ annealing temperature. The highest high temperature ductility was 19% of elongation and 65% of reduction in area. (3) The fracture toughness in this experiment was above 90 MPa m1/2 and the fracture toughness of ␤ annealing was slightly higher than those of ␤ processing. The results showed that fracture toughness was more sensitive to microstructure type. Besides, grain boundary ␣, width of lamellar ␣ and ␣ colony size also had significant effect on fracture toughness. (4) SEM observation showed that the fracture mechanism at tensile test changed from transcrystalline in the ␤ processing conditions to a mixture of intercrystalline and transcrystalline at the ␤ annealing conditions. The situation also applied to the fracture toughness test. While at high temperature tensile test both of the two conditions were typical ductile dimple fracture pattern.

63

Acknowledgements The authors would like to thank Mr. Y.Q. Ning, Z.F. Shi, C. Qin, Z.L. Zhao for useful technical discussions and extensive experiments during the finish of this project.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

W.-S Lee, C.-F. Lin, Mater. Sci. Eng. A 241 (1998) 48–59. B.D. Venkatesh, et al., Mater. Sci. Eng. A 506 (2009) 117–124. R. Ding, et al., Mater. Sci. Eng. A 327 (2002) 233–245. T. Seshacaryulu, et al., Mater. Sci. Eng. A 279 (2000) 289–299. X.N. Wang, L. Tong, Z.S. Zhu, et al., Rare Metal Mater. Eng. 37 (2008) 537–539. M.N. Gungor, et al., Mater. Sci. Eng. A 410–411 (2005) 369–374. R. Filip, K. Kubiak, W. Ziaja, et al., J. Mater. Process. Technol. 133 (2003) 84–89. S.L. Serniatin, G.D. Lahoti, Metall. Trans 12A (1981) 1705. J.J. Jonas, M.J. Luton, Advances in Deformation Processing, Plenum Press, New York, 1978, p. 215. C.M. Young, O.D. Sherby, in: A.L. Hoffmanner (Ed.), Metal Forming-Interrelation between Theory and Practice, Plenum Press, New York, 1971, p. 429. Y. Liu, T.N. Baker, Mater. Sci. Eng. A 197 (1995) 125–131. I. Weiss, F.H. Froes, D. Eylon, G.E. Welsch, Metall. Trans. 17A (1986) 1935. G. Lutjering, J.C. Williams, Titanium, Berlin Heidelberg, NewYork, 2007. S. Ankem, H. Margolin, Metall. Trans. A 17 (1986) 2209–2226. B.A. Movchan, F.D. Lemkey, Mater. Sci. Eng. A 224 (1997) 36–45. S. Ankem, Ph.D Thesis, Polytechnic Institute of New York, 1980. T. Wang, H.Z. Guo, Y.W. Wang, X.N. Peng, et al., Mater. Sci. Eng. A 528 (2011) 2370–2379. Y.Q. Ning, Z.K. Yao, H. Li, H.Z. Guo, T. Yu, Y.W. Zhang, Mater. Sci. Eng. A 527 (2010) 961–966. C. Sauer, G. Lutjering, Mater. Sci. Eng. A 319–321 (2001) 393–397. K. Keshava Murthy, S. Sundaresan, Eng. Fract. Mech. 58 (1–2) (1997) 29–41. M. Niinomi, T. Kobayashi, Mater. Sci. Eng. A 213 (1996) 16–24. M.A. Greenfield, H. Margolin, Metall. Trans. 2 (3) (1971) 841–847.