- Email: [email protected]
Designed bimodal size lamellar O microstructures in Ti2AlNb based alloy: Microstructural evolution, tensile and creep properties
Materials Science & Engineering A 618 (2014) 288–294
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Designed bimodal size lamellar O microstructures in Ti2AlNb based alloy: Microstructural evolution, tensile and creep properties Wang Wei a,n, Zeng Weidong a, Xue Chen a, Liang Xiaobo b, Zhang Jianwei b a b
State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi’an 710072, China BeiJing Iron & Steel Research Institute, Beijing 100081, China
art ic l e i nf o
a b s t r a c t
Article history: Received 30 May 2014 Received in revised form 30 August 2014 Accepted 5 September 2014 Available online 16 September 2014
Microstructure evolution, tensile and creep properties of designing bimodal size lamellar O phases by thermo-mechanical processing including conventional forging, isothermal forging process and heat treatment for Ti–22Al–25Nb (at%) orthorhombic alloy were investigated. The microstructures were obtained by different solution- and age-treatment temperatures, and analyzed by the BSE technique. The creep behavior of the alloy was studied at 650 1C/150 MPa for 100 h in air. The tensile strength of the alloy at room temperature and 650 1C was also investigated. The experimental results showed that the microstructure of the isothermally forged alloys at 1080 1C contained the non-uniform distribution of lamellar O phases and B2 matrix. The advantage of the bimodal lamellar size distributed alloy can be concluded that firstly, the coarse lamellar O formed during solution process makes the alloy owns good elongation and secondly, the fine lamellar O precipitate during the aging process strengthens the alloy. The volume fraction and mean thickness of the lamellar O could be well controlled by the heat treatment. The yield strength was sensitive to the thickness of lamellar O, increase in the aging temperature leads to a decrease in strength and an increase in ductility. The relationship between creep resistance of alloys and microstructural features, such as the volume fraction of each phases, the morphology of lamellar O phases was also discussed. & 2014 Elsevier B.V. All rights reserved.
Keywords: Ti2AlNb-based alloys Orthorhombic Microstructural evolution Creep Tensile
1. Introduction Recently, a new class of titanium intermetallic alloys, based on the orthorhombic Ti2AlNb phase, has been receiving attention as potential materials for aircraft engine applications due to their high strength-toweight ratio, greater fracture toughness, and better workability than conventional intermetallic alloys such as TiAl-based and Ti3Al-based alloys [1–4]. A typical orthorhombic alloy is Ti–22Al–25Nb (at%), in which a high-temperature B2 phase was incorporated in the alloy's constitution to further improve ductility and fracture toughness. The alloys, which primarily consists of an OþB2 two-phase microstructure, is said to have the optimum combination of strength, creep, and fracture toughness properties [5–7]. The microstructure of Ti2AlNb based alloys can be varied in a wide range depending on processing methods and subsequent heat treatments [8]. Similar to conventional titanium alloys, equiaxed, and duplex, near lamellar and fully lamellar microstructures can be obtained by appropriate thermo-mechanical processing (TMP) and heat treatment [9]. A duplex microstructure of Ti2AlNb based alloys shows good ductility and strength at both low and high temperatures, but poor fracture toughness and creep resistance. In contrast, a fully n
Corresponding author. Tel.: þ 86 29 88494298. E-mail addresses: [email protected] (W. Wei), [email protected] (Z. Weidong). http://dx.doi.org/10.1016/j.msea.2014.09.035 0921-5093/& 2014 Elsevier B.V. All rights reserved.
lamellar microstructure provides good creep resistance and fracture toughness, but shows low ductility at room temperature [10]. Especially, the high creep resistance makes these alloys very interesting for high temperature structural parts. For practical application of Ti2AlNb alloys, their ductility, toughness, and strength have to be optimized by means of appropriate microstructural design. The microstructural design requires detailed knowledge of the relationship between microstructure and mechanical properties. As a compromise, so-called designed bimodal size lamellar O microstructures were proposed which offers a good balance between the mechanical properties of both the duplex and the fully lamellar microstructure [11]. In the present work, the relationship between TMP and the mechanical property for Ti–22Al–25Nb alloys was studied. The parameters of TMP including isothermal forging, solution- and agetreatment temperature for obtaining better mechanical properties were identified, and the microstructural characteristics for different O phases were revealed. This information will be beneficial to practical production of high performance Ti–22Al–25Nb alloys.
2. Materials and experiments The Ti–22Al–25Nb bar was provided by Central Iron and Steel Research Institute (CISRI) with a dimension of Φ240 360 mm. The chemical analysis is shown in Table 1, it can be seen that the
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
bar composition was in agreement with nominal composition, and the gas impurity contents were lower (e.g., oxygen r500 ppm, hydrogen r300 ppm and nitrogen r800 ppm ). The microstructure of the Ti–22Al–25Nb bar can be seen in Fig. 1(a). The microstructure contained B2 matrix, equiaxed α2/O particles, and lamellar O. The equiaxed α2/O particles consisted of α2 particle and rim-O (Fig. 1(b)) which was formed by the peritectoid reaction of α2 phase and B2 matrix. The pinning effect of the equiaxed particles is beneficial for restraining the B2 grain growth. The B2 grain size of the Ti–22Al–25Nb bar after eight deformation times is measured to be 38 μm in averaged. The lamellar O (Fig. 1(b)) was precipitated during cooling process after forging. In order to obtain fully lamellar O phases and enhance the mechanical properties of alloys, the isothermal forging processing in 1080 1C (B2 phase region) is adopted. The heat-treatment experimental specimens of 14 mm 14 mm 12 mm were machined along the compression axis on the low-speed wire electrical discharge machine (WEDM). In order to investigate microstructural evolution of alloys, the specimens were solution-treated between 920 1C and 1000 1C at intervals of 20 1C for 1 h followed by water quenching (WQ), and aged between 760 1C and 840 1C at intervals of 20 1C for 12 h followed by air cooling (AC). The alloys after isothermal forging were solution-treated at 940 1C for 1 h followed by water quenching (WQ) and then age-treated at 760 1C, 800 1C and 840 1C for 12 h by air cooling (AC) to study the creep and tensile properties. To investigate the microstructural evolution, the samples isothermal forged, solution- and age-treated were characterized using the BSE (back scattered electron) mode in SEM (scanning electron microscope). The specimens for SEM were prepared by standard metallographic techniques. The phase structural analysis was performed by X-ray diffraction (XRD). Diffraction utilized Cu Kα radiation in the angular range of 2θ ¼20–901 at a recording rate of 2 1/min. The finer microstructure details of the alloys before and after creep deformation were carried out in a JEM-200CX transmission electron microscopy (TEM). The TEM samples were prepared by the following steps. Firstly, the samples were cut from the alloys after heat treatment state and creep deformation. Then, they were thinned down to 30–50 μm by polishing with SiC paper. Finally, they were punched to 3 mm disc samples. The electrochemical polishing is the final stage of thinning in a solution of 6% perchloric
289
acid, 34% n-butanol and 60% carbinol at 30 1C with an applied current of 70 mA and a voltage of 30 V. Fig. 2(a) shows the cylindrical tensile specimens. Fig. 2 (b) shows the cylindrical creep specimens. The creep specimens with a gauge diameter of 5 mm and length of 70 mm were machined after heat treatment. Creep testing was conducted in air using RD testing machine. The testing temperature was 650 1C. The applied stress was 150 MPa. The creep time was 100 h. The temperature fluctuations were less than 72 1C. The tensile samples were 12 mm in diameter and 70 mm in gauge length. The tensile tests were carried out in air at room temperature using an Instron 1185 mechanical testing machine. The high temperature tension experiments were performed at 650 1C using a MTS mechanical testing machine.
3. Results and discussion 3.1. Investigation of bimodal lamellar size distribution The microstructure of the isothermally forged alloys at 1080 1C is shown in Fig. 3. It can be seen from the XRD analysis results (Fig. 3 (a)), the alloys contained O phase and B2 matrix. In BSD image, the black equiaxed particles are α2 phases, the gray regions are O phases, and light regions are B2 phases. Lamellar O phases and B2 matrix could be seen in Fig. 3(b). A very small amount of lamellar α2 phases are also presented in Fig. 3(b). The diffraction peak of the α2 phase is very weak, so it is very difficult to detect the α2 phases in Fig. 3(a). The α2 and lamellar O-phases were formed during the gradual air
Table 1 Chemical composition of the Ti–22Al–25Nb alloy (at%) (the true chemical composition of the alloy was measured with a chemical analysis method, which was Ti–10.8Al–43.0Nb–0.050O–0.007C–0.005N–0.004H in weight percent (wt%)). Ti
Al
Nb
O
N
H
Bal
22.3
25.7
0.00043
0.000052
0.000009 Fig. 2. Geometry and dimensions of tensile (a) and creep (b) testing specimens
Fig. 1. Microstructure of the as-forged alloy (a) macro-microstructure; (b) microstructure at higher magnification ( the black equiaxed particles are α2 phases, the gray regions are O phases, and light regions are B2 phases).
290
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
cooling to room temperature after isothermally forging. In comparison with the as-forged alloy, more lamellar O phases precipitated and equiaxed α2 particles transformed into O or B2 matrix. The initial B2 grain size of the alloys produced by non-isothermal forging is well controlled and counted to be 37 μm–72 μm due to the pinning effect of the α2 particles; while the average B2 grain size of the alloys after isothermal forging at 1080 1C is about 320 μm. During heating and isothermal forging process at 1080 1C, the equiaxed α2 particles were gradually decreasing and go away completely. So the B2 grain size is greater in microstructures. Although the lamellar O phases could be obtained in isothermal forging processing at 1080 1C, the distribution of the lamellar O phases is inhomogeneous and the lamellar O phases are very fine. In order to obtain bimodal size lamellar O microstructures, the microstructures after isothermally forging at 1080 1C were solution- and age-treated in α2 þB2þ O and OþB2 phase regions. The microstructures and phases of Ti–22Al–25Nb alloy after water quenching from different solution temperatures are shown in Fig. 4(a) illustrates the microstructure of the isothermally forged Ti–22Al–25Nb alloy solution treated at 920 1C. It mainly contained lamellar O phases and B2 matrix. In comparison
with the as-isothermal forged alloy, the very fine-lamellar O phases were dissolved, the coarser-lamellar O phases formed and became wider and shorter, and some have already become globular during the heat-treatment. As the solution temperature increased to 960 1C (Fig. 4 (b)), more lamellar O phases dissolved into B2. The volume fraction of lamellar O decreased, and the shape became shorter and coarser. The amount of the O phase precipitates decreased obviously as the solution temperature increases to 980 1C which can be seen in Fig. 4 (c). This microstructure which is clearly different from the microstructures below 980 1C contains lamellar α2 /O and B2. Note that α2 phase formed as acicular in the lamellar O phase in 980 1C (this temperature is located in α2 þ B2þO phase field). This α2 phase has an orientation relationship to the parent O phase [12], ½001O ==½0001α2 and (110)o/ /(101̄ 0Þα2 [12]. The microstructure is very sensitive to the solution treatment temperature. When the alloys solution-treated in different phase region, the microstructures are different. The coarse-lamellar O phases were obtained in solution-treatment process. In order to realize the bimodal size lamellar O microstructures, aging treatments were performed on selected solution-treated samples. Fig. 5(a)–(e) illustrates the
Fig. 3. Phase and microstructure of the isothermally forged alloys at 1080 1C: (a) XRD image (b) SEM image (the black equiaxed particle is α2 phase, the gray is O phase and light regions are B2).
Fig. 4. Microstructures of solution-treated and water-quenched specimens. Solution-treatment temperatures were (a) 920 1C, (b) 960 1C and (c) 980 1C.
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
microstructures solution-treated at 940 1C and then age-treated between 760 1C and 840 1C for 12 h, followed by air cooling. All the aged microstructures were comprised of secondary acicular precipitates of the O phase within the B2 grains. The reason is believed to be the low-energy configuration for specific sets of planes in the O and B2 phases (i.e., OR) [13]. The size and volume fraction of the acicular precipitates of the O phase depend on the aging temperature. The largest precipitate size occurred at 840 1C and the finest size was displayed at 760 1C. It also can be seen from the figures that the amount of the secondary acicular O precipitates decreases as the age treatment temperature increases. From Fig. 5, noted that the more very fine acicular α2 phases are presented in lower age-treatment temperature (760 and 780 1C). As the temperature increased to 800– 840 1C, the amount of the fine acicular α2 phases is very rare and the width of α2 phases became thicker. Although the samples were agetreated in B2þO phase region for 12 h, the precipitation occurred with a lenticular morphology, and the primary lamellar O remained relatively unchanged. This is important for the design of the bimodal size lamellar O microstructures because the primary coarse-lamellar O phases and volume fraction can be controlled by selecting the appropriate solution temperatures and the secondary acicular precipitate size and volume fraction can be controlled by selecting the appropriate aging temperatures. According to the previous researches [8], the primary coarse-lamellar O make a greater contribution for the creep resistance, but exhibits poor strengthening effect compares to the fine secondary acicular O. Thus, to better adjust these microstructure parameters is believed to be important for obtaining the superior mechanical properties of the alloy.
291
and 10.2% for HT-760, 1116 MPa, 975 MPa, 10.5% and 11% for HT800, and 1066 MPa, 926 MPa, 12.5% and 12% for HT-840. Thus, the tensile strength of HT-760 was greater than that for HT-840; however it exhibits poorer process ability and lower elongation values. The various parameters which influence the yield strength are (1) the grain size of the B2 phase, (2) the strength of the individual phases as influenced by changes in their composition with heattreatment, (3) the volume fractions of α2/O and B2 phases, and (4) the lath size of O phase, (5) dislocation substructure [14]. No estimate has been made of the variation of the first factors in this study. The heat-treatment regimes of HT-760, HT-800, and HT-840 are the same except for the difference of age-treatment. The grain size of the B2 phase does not vary with age-treatment. Since the α2 phase volume fraction is extremely little for three microstructures and the tensile properties of Ti–22Al–25Nb alloy are significantly influenced by the microstructures, especially the volume fraction of B2 and O phases, and size of secondary lamellar O phase. From Table 2, it can be seen that increase in aging temperature leads to a decrease in strength and an increase in ductility. This variation trend of strength is caused by the noticeable coarsening besides the volume fraction reduction of secondary lamellar O phase with increasing aging temperatures [15]. The result is in agreement with literature [16], the yield strength is associated with the size and morphology of constituent phases. The yield strength can be expressed by the following rule of mixture:
σ a ¼ σ O f O þ σ 0B2 f B2 where σa is the yield strength of alloy; fo and fB2 designate the volume fraction of primary O phase and B2 phase, respectively; σo is the yield strength of O phase; σ B2' is a complex parameter that is determined by the size of B2 after thermo-mechanical processing and the volume fraction of O phase evolved in primary B2 matrix during heat treatment followed. For a given solutiontreatment temperature, the change of age-treatment temperature can influence the volume fraction of B2 and O phase, and the geometrical features of O phases.
3.2. Tensile properties The room temperature (RT) and 650 1C high temperature tensile properties, including Yield strength (YS), Ultimate tensile strength (UTS), Elongation (E), and Reduction area (RA), for selected microstructures are listed in Table 2. In most cases the results represent the average of at least three tests. The UTS, YS, E and RA of RT tensile properties were 1123 MPa, 997 MPa, 10%
Table 2 Tensile properties of the Ti–22Al–25Nb alloy (at%) at different heat-treatment temperatures. Testing temperature/1C
Sample
ST Temperature/1C
AT Temperature/1C
Ultimate tensile strength/MPa
Yield strength/MPa
Elongation/%
Reduction area/%
RT
HT-760 HT-800 HT-840
940 940 940
760 800 840
1123 1116 1066
997 975 926
10 10.5 12.5
10.2 11 12
650
HT-760 HT-800 HT-840
940 940 940
760 800 840
937 906 880
845 830 780
14 15 15.5
59 59 54
ST¼solution treatment and AT ¼ age treatment. Table 3 Heat treatment schedules and the corresponding microstructural parameters of the Ti–22Al–25Nb alloy. Vp, Llath, Wlath, Lacicular, Wacicular and d* are the volume fraction of each phase, the length of primary coarse-lamellar O, the width of primary coarse-lamellar O, the length of secondary fine acicular O, the width of secondary fine acicular O, and the grain size of B2 phase, respectively. ε is the average errors. Sample
α2 Vp ( 7 ε)
O Vp ( 7ε)
B2 Vp ( 7ε)
dn (μm) ( 7ε)
Llath /μm ( 7 ε)
Wlath /μm ( 7 ε)
Lacicular /μm ( 7ε)
Wacicular /μm ( 7 ε)
HT-760
1.85 (0.05)
72.98 (2.2)
25.17 (1.3)
217 (5.4)
4.04 (0.2)
0.28 (0.02)
0.91 (0.04)
0.09 (0.003)
HT-800
0.97 (0.04)
67.68 (1.8)
31.35 (1.3)
219 (6.1)
3.83 (0.4)
0.37 (0.03)
1.47 (0.07)
0.16 (0.004)
HT-840
0.31 (0.02)
63.62 (2.3)
36.07 (1.1)
220 (5.8)
2.15 (0.3)
0.42 (0.04)
2.13 (0.07)
0.23 (0.004)
292
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
From Table 3, it can be seen that the volume fraction of O phase, B2 phase and α2 phase are 72.98%, 25.17%, and 1.85% for HT-760; 67.68%, 31.35%, and 0.97% for HT-800; and 63.62%, 36.07%, and 0.31% for HT-840. It also can be seen that as the age temperature increased, the length and width of primary coarse-lamellar O became shorter and coarser, while the length and width of secondary acicular O became longer and coarser. The volume fraction of B2 for HT-760 is lowest than that for others and the volume fraction of O for HT-760 is highest, while the YS and UTS at RT of HT-760 outperformed HT-840. In the present work, tensile samples were solution treated at a constant temperature, so the volume fraction and the size of primary lamellar O are constant. The volume fraction of secondary acicular O phases increases as the aging decreases for the same solution temperatures. It means that the strength of alloy is mainly about precipitation strengthening of secondary acicular O-phases. The E values were 10% for the HT-760 RT tensile experiments, while the E values were 12.5% for the HT-840 RT tensile experiments. This difference of E values is mainly ascribed to the volume fraction of the O phase and B2 phase. O-phase platelets also provided a significant strengthening effect. Although the ordered O-phase is stronger than the B2-phase, it suffers from a lower number of available slip systems. Compared with the B2 structure, the O-phase suffers from a lower number of available slip systems [17]. Even though the
active slip systems between the O and BCC phases are compatible, the limited number of O-phase slip systems is considered to be responsible for the low E values in O-phase dominated microstructures. So the E values for the HT-840 RT outperformed the E values for the HT-760 RT. The UTS for HT-760 outperformed the UTS for HT-840, the higher O-phase volume fractions and finer O-phase laths were expected to be the probable reason for this behavior. However, Ophase dominated polycrystalline microstructures have proven to be brittle [3,16]. Thus, maintaining adequate volume fractions of the BCC phase is necessary for providing ductility. The fractographs of the room-temperature tensile specimens are shown in Fig. 6. It can be seen that the fracture mechanism of the material is a mixture of cleavage and faceted fracture. In addition, the specimens exhibited a flat fracture characteristic of brittle metals. The microstructures of alloys exhibited a combination of the fracture characteristics: both cleaved and faceted O grains and dimples within B2 regions. For HT-760, planar-slip traces were identified in the O phase, O/O cracking developed, while for HT-840, wavy slip was exhibited in B2 phase, slip compatibility between O and B2 grains was observed. The 650 1C tensile results are shown in Table 2. The tensile specimen for HT-760, which exhibited a YS¼ 845 MPa and a UTS¼937MPa, was significantly stronger than the tensile specimen for HT-840, which exhibited a tensile YS ¼780 MPa and a
Fig. 5. Comparison of a 940 1C/1 h solution-treated specimen and aged at different temperatures. The specimens were aged at (a) 760 1C, (b) 780 1C, (c) 800 1C, (d) 820 1C, and (e) 840 1C for 12 h followed by air cooling.
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
UTS¼ 880 MPa. For all temperatures, the YS decreased with increasing temperature for the alloys. The highest 650 1C YS value was exhibited in HT-760, greater O-phase concentrations tended to provide greater elevated-temperature strength. This suggests that the O phase provides greater elevated-temperature precipitation strengthening in these alloys.
293
coarser lamellar O and secondary acicular lamellar O phases became coarser. According to literature [2], the increased thickness of lamellar O-phase is beneficial to the improvement of the creep resistance of alloys. However, the facts are to the contrary. We
3.3. Creep property The creep behavior of the alloy for three heat treatment schedules has also been studied. Fig. 7 shows the creep curves of the three microstructures at 150 MPa and 650 1C for 100 h. Each plot possesses the typical primary or transient creep stage followed by a steady-state stage. No obvious tertiary creep stage was observed. The shape of creep curves is essentially the same, independently of the three microstructural variables. It can be seen that the creep rate of the primary creep stage for HT-840 is higher than that of HT-800 and HT-760. HT-760 has best creep resistance at 650 1C/150 MPa as compared with others. It mainly relates to the microstructures. From Table 3, the volume fraction of α2 phase is much less, the most volume fraction of α2 phase for HT-760 is only 1.85%, so we can probably ignore the impact of α2 phase on mechanical properties. For the microstructure of the HT-760 and the HT-840, as the age temperature increased, the width of primary
Fig. 7. The creep curves of the three heat-treatment schedules at 150 MPa and 650 1C for 100 h.
Fig. 6. Room-temperature fractographs for the samples: (a) macro-fractograph for samples HT-760; (b), (c) micro-fractograph for samples HT-760; (e) macro-fractograph for samples HT-840; (f), and (g) micro-fractograph for samples HT-840.
294
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
Fig. 8. TEM images of creep sample at 650 1C/150 MPa (a) HT-840 and (b) HT-760.
conjecture that except for the lamellar O phase, the volume fraction of B2 phase can also significantly influence the creep properties. For the three microstructures, the volume fraction of B2 for HT-840 is higher than that of HT-760, while the volume fraction of O for HT840 is lower than that of HT-760. The lamellar O phases have better creep resistance than B2 phases and the volume fraction of secondary acicular lamellar O phases for HT-760 is higher than that of HT-800 and HT-840, so HT-760 has the best creep resistance. In order to further study the effect of microstructure on creep behavior, the bright-field TEM images of the creep sample HT-760 and HT-840 have been present in Fig. 8. From Fig. 8, it can be seen that the creep resistance of coarser- and fine-lamellar O phase is different. The density of dislocations in the fine-lamellar O-phase seems to be very high, while the density of dislocations in the coarser-lamellar O-phase is low. It means the movement of dislocations in coarser-lamellar O-phase is not so active during creep deformation at 650 1C; so the creep resistance of coarserlamellar O-phase is higher than that of fine-lamellar O-phase. Fig. 8(a) shows the dislocations observed in crept sample HT-840. The microstructure is comprised of coarser- and fine-lamellar O phase, and B2 matrix. Because of more B2 phases present in this microstructure, the coarser- lamellar O was bent. From the microstructure for HT-760 (Fig. 8(b)), tangled or bowing dislocations and any kink or jog are not observed. Only few straight dislocations can be observed in the crept alloy. It means the movement of dislocations in this condition is not so active during creep deformation. No microstructural observations were recorded to suggest that grain boundary sliding contributes to the creep strains in this study. So, it was therefore suggested that the creep mechanism of HT-760 are dominated by a dislocation-controlled creep process, and this conclusion is supported by the observation of the dislocation densities. The current observations were insufficient to determine whether a glide or climb mechanism was rate controlling. 4. Conclusion In this study, microstructure evolution and tensile and creep properties of designing bimodal size lamellar O phases by thermomechanical processing including common forging, isothermal forging process and heat treatment for Ti–22Al–25Nb (at%) orthorhombic alloy were investigated. The effect of age-treatment temperature on microstructure and tensile and creep properties are also investigated. The microstructure of the isothermally forged alloys at 1080 1C contained the non-uniform distribution of lamellar O phases and B2 matrix. The lamellar O could be obtained when solution treated at OþB2 phase region and the secondary acicular O was emerged during the aging process. The volume fraction and thickness of coarse and acicular lamellar O could be well controlled by the solution and aging treatment. During aging treatments, the size and volume fraction of secondary acicular lamellar O are very
sensitive to temperature. The volume fraction decreased and the size increased with increasing of aging temperature. The size of secondary acicular lamellar O phase in α2 þ OþB2 phase region solution treatment plus aging condition is larger than that in OþB2 phase region solution treatment plus aging treatment. The size and volume fraction of secondary acicular lamellar O affect the mechanical properties of the alloy strongly. In terms of the RT tensile behavior, the YS and UTS at RT of HT-760 outperformed HT-840; however the E and RA of HT-760 have a lower value. The trends of 650 1C tensile properties were similar to those at RT; however, each of the samples exhibited lower strength and higher E and RA values than at RT. In terms of the creep behavior, HT-760 has best creep resistance at 650 1C/150 MPa as compared with HT-800 and HT-840. The creep properties of alloy were concerned with the microstructures. The volume fraction of B2 for HT-840 is higher than that of HT-760, while the volume fraction of O for HT-840 is lower than that of HT-760. The lamellar O phases have better creep resistance than B2 phases, so HT-760 has best creep resistance.
Acknowledgments This work was financially supported by Research Fund for the Doctoral Program of Higher Education of China with No. 201161 02110015, the New Century Excellent Talents in University with No. NCET-07-0696, and the National 973 Project of China with No. 2007CB613807. References [1] D. Banerjee, A.K. Gogia, T.K. Nandy, V.A. Joshi, Acta Metall. 36 (1988) 871. [2] A.K. Gogia, T.K. Nandy, D. Banerjee, T. Carisey, J.L. Strudel, J.M. Franchet, Intermetallics 6 (1998) 741. [3] C.J. Boehlert, D.B. Miracle, Metall. Mater. Trans. A 30 (1999) 2349–2367. [4] M. Hagiwara, T. Kitaura, Y. Ono, T. Yuri, T. Ogata, S. Emura, Mater. Trans. 53 (2012) 1138–1147. [5] Chen Xue, Weidong Zeng, B.i.n. Xu, Xiaobo Liang, Jianwei Zhang, Shiqiong Li, Intermetallics 29 (2012) 41–47. [6] L. Germann, D. Banerjee, J.Y. Guédou, J.-L. Strudel, Intermetallics 13 (2005) 920–924. [7] Wei Wang, Weidong Zeng, Chen Xue, Xiaobo Liang, Jianwei Zhang, Intermetallics 45 (2014) 29–37. [8] C.J. Cowen, C.J. Boehlert, Intermetallics 14 (2006) 412–422. [9] Jörg Kumpfert, Adv. Eng. Mater. 11 (2001) 851–864. [10] Y.W. Kim, D.M. Dimiduk, J. Manag. 43 (1991) 40–47. [11] Chen Xue, Weidong Zeng, Wei Wang, Xiaobo Liang, Jianwei Zhang, Mater. Sci. Eng. A 587 (2013) 54–60. [12] K. Muraleedharan, D. Banerjee, S. Banerjee, S. Lele, Philos. Mag. 71 (1995) 1011–1036. [13] C.J. Boehlert, B.S. Majumdar, V. Seetharaman, D.B. Miracle, Metall. Mater. Trans. A 30 (1999) 2305–2323. [14] A.K. Gogia, T.K. Nandy, K. Muraleedharan, D. Banerjee, Mater. Sci. Eng. A 159 (1992) 73–86. [15] C.L. Li, X.J. Mi, W.J. Ye, S.X. Hui, Y. Yu, W.Q. Wang, J. Alloy Compd. 550 (2013) 23–30. [16] C.J. Boehlert, D.B. Miracle, Metall. Mater. Trans. A 38 (2007) 26–34. [17] C.J. Boehlert, Metall. Mater. Trans. A 32 (2001) 1977–1988.
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Designed bimodal size lamellar O microstructures in Ti2AlNb based alloy: Microstructural evolution, tensile and creep properties Wang Wei a,n, Zeng Weidong a, Xue Chen a, Liang Xiaobo b, Zhang Jianwei b a b
State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi’an 710072, China BeiJing Iron & Steel Research Institute, Beijing 100081, China
art ic l e i nf o
a b s t r a c t
Article history: Received 30 May 2014 Received in revised form 30 August 2014 Accepted 5 September 2014 Available online 16 September 2014
Microstructure evolution, tensile and creep properties of designing bimodal size lamellar O phases by thermo-mechanical processing including conventional forging, isothermal forging process and heat treatment for Ti–22Al–25Nb (at%) orthorhombic alloy were investigated. The microstructures were obtained by different solution- and age-treatment temperatures, and analyzed by the BSE technique. The creep behavior of the alloy was studied at 650 1C/150 MPa for 100 h in air. The tensile strength of the alloy at room temperature and 650 1C was also investigated. The experimental results showed that the microstructure of the isothermally forged alloys at 1080 1C contained the non-uniform distribution of lamellar O phases and B2 matrix. The advantage of the bimodal lamellar size distributed alloy can be concluded that firstly, the coarse lamellar O formed during solution process makes the alloy owns good elongation and secondly, the fine lamellar O precipitate during the aging process strengthens the alloy. The volume fraction and mean thickness of the lamellar O could be well controlled by the heat treatment. The yield strength was sensitive to the thickness of lamellar O, increase in the aging temperature leads to a decrease in strength and an increase in ductility. The relationship between creep resistance of alloys and microstructural features, such as the volume fraction of each phases, the morphology of lamellar O phases was also discussed. & 2014 Elsevier B.V. All rights reserved.
Keywords: Ti2AlNb-based alloys Orthorhombic Microstructural evolution Creep Tensile
1. Introduction Recently, a new class of titanium intermetallic alloys, based on the orthorhombic Ti2AlNb phase, has been receiving attention as potential materials for aircraft engine applications due to their high strength-toweight ratio, greater fracture toughness, and better workability than conventional intermetallic alloys such as TiAl-based and Ti3Al-based alloys [1–4]. A typical orthorhombic alloy is Ti–22Al–25Nb (at%), in which a high-temperature B2 phase was incorporated in the alloy's constitution to further improve ductility and fracture toughness. The alloys, which primarily consists of an OþB2 two-phase microstructure, is said to have the optimum combination of strength, creep, and fracture toughness properties [5–7]. The microstructure of Ti2AlNb based alloys can be varied in a wide range depending on processing methods and subsequent heat treatments [8]. Similar to conventional titanium alloys, equiaxed, and duplex, near lamellar and fully lamellar microstructures can be obtained by appropriate thermo-mechanical processing (TMP) and heat treatment [9]. A duplex microstructure of Ti2AlNb based alloys shows good ductility and strength at both low and high temperatures, but poor fracture toughness and creep resistance. In contrast, a fully n
Corresponding author. Tel.: þ 86 29 88494298. E-mail addresses: [email protected] (W. Wei), [email protected] (Z. Weidong). http://dx.doi.org/10.1016/j.msea.2014.09.035 0921-5093/& 2014 Elsevier B.V. All rights reserved.
lamellar microstructure provides good creep resistance and fracture toughness, but shows low ductility at room temperature [10]. Especially, the high creep resistance makes these alloys very interesting for high temperature structural parts. For practical application of Ti2AlNb alloys, their ductility, toughness, and strength have to be optimized by means of appropriate microstructural design. The microstructural design requires detailed knowledge of the relationship between microstructure and mechanical properties. As a compromise, so-called designed bimodal size lamellar O microstructures were proposed which offers a good balance between the mechanical properties of both the duplex and the fully lamellar microstructure [11]. In the present work, the relationship between TMP and the mechanical property for Ti–22Al–25Nb alloys was studied. The parameters of TMP including isothermal forging, solution- and agetreatment temperature for obtaining better mechanical properties were identified, and the microstructural characteristics for different O phases were revealed. This information will be beneficial to practical production of high performance Ti–22Al–25Nb alloys.
2. Materials and experiments The Ti–22Al–25Nb bar was provided by Central Iron and Steel Research Institute (CISRI) with a dimension of Φ240 360 mm. The chemical analysis is shown in Table 1, it can be seen that the
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
bar composition was in agreement with nominal composition, and the gas impurity contents were lower (e.g., oxygen r500 ppm, hydrogen r300 ppm and nitrogen r800 ppm ). The microstructure of the Ti–22Al–25Nb bar can be seen in Fig. 1(a). The microstructure contained B2 matrix, equiaxed α2/O particles, and lamellar O. The equiaxed α2/O particles consisted of α2 particle and rim-O (Fig. 1(b)) which was formed by the peritectoid reaction of α2 phase and B2 matrix. The pinning effect of the equiaxed particles is beneficial for restraining the B2 grain growth. The B2 grain size of the Ti–22Al–25Nb bar after eight deformation times is measured to be 38 μm in averaged. The lamellar O (Fig. 1(b)) was precipitated during cooling process after forging. In order to obtain fully lamellar O phases and enhance the mechanical properties of alloys, the isothermal forging processing in 1080 1C (B2 phase region) is adopted. The heat-treatment experimental specimens of 14 mm 14 mm 12 mm were machined along the compression axis on the low-speed wire electrical discharge machine (WEDM). In order to investigate microstructural evolution of alloys, the specimens were solution-treated between 920 1C and 1000 1C at intervals of 20 1C for 1 h followed by water quenching (WQ), and aged between 760 1C and 840 1C at intervals of 20 1C for 12 h followed by air cooling (AC). The alloys after isothermal forging were solution-treated at 940 1C for 1 h followed by water quenching (WQ) and then age-treated at 760 1C, 800 1C and 840 1C for 12 h by air cooling (AC) to study the creep and tensile properties. To investigate the microstructural evolution, the samples isothermal forged, solution- and age-treated were characterized using the BSE (back scattered electron) mode in SEM (scanning electron microscope). The specimens for SEM were prepared by standard metallographic techniques. The phase structural analysis was performed by X-ray diffraction (XRD). Diffraction utilized Cu Kα radiation in the angular range of 2θ ¼20–901 at a recording rate of 2 1/min. The finer microstructure details of the alloys before and after creep deformation were carried out in a JEM-200CX transmission electron microscopy (TEM). The TEM samples were prepared by the following steps. Firstly, the samples were cut from the alloys after heat treatment state and creep deformation. Then, they were thinned down to 30–50 μm by polishing with SiC paper. Finally, they were punched to 3 mm disc samples. The electrochemical polishing is the final stage of thinning in a solution of 6% perchloric
289
acid, 34% n-butanol and 60% carbinol at 30 1C with an applied current of 70 mA and a voltage of 30 V. Fig. 2(a) shows the cylindrical tensile specimens. Fig. 2 (b) shows the cylindrical creep specimens. The creep specimens with a gauge diameter of 5 mm and length of 70 mm were machined after heat treatment. Creep testing was conducted in air using RD testing machine. The testing temperature was 650 1C. The applied stress was 150 MPa. The creep time was 100 h. The temperature fluctuations were less than 72 1C. The tensile samples were 12 mm in diameter and 70 mm in gauge length. The tensile tests were carried out in air at room temperature using an Instron 1185 mechanical testing machine. The high temperature tension experiments were performed at 650 1C using a MTS mechanical testing machine.
3. Results and discussion 3.1. Investigation of bimodal lamellar size distribution The microstructure of the isothermally forged alloys at 1080 1C is shown in Fig. 3. It can be seen from the XRD analysis results (Fig. 3 (a)), the alloys contained O phase and B2 matrix. In BSD image, the black equiaxed particles are α2 phases, the gray regions are O phases, and light regions are B2 phases. Lamellar O phases and B2 matrix could be seen in Fig. 3(b). A very small amount of lamellar α2 phases are also presented in Fig. 3(b). The diffraction peak of the α2 phase is very weak, so it is very difficult to detect the α2 phases in Fig. 3(a). The α2 and lamellar O-phases were formed during the gradual air
Table 1 Chemical composition of the Ti–22Al–25Nb alloy (at%) (the true chemical composition of the alloy was measured with a chemical analysis method, which was Ti–10.8Al–43.0Nb–0.050O–0.007C–0.005N–0.004H in weight percent (wt%)). Ti
Al
Nb
O
N
H
Bal
22.3
25.7
0.00043
0.000052
0.000009 Fig. 2. Geometry and dimensions of tensile (a) and creep (b) testing specimens
Fig. 1. Microstructure of the as-forged alloy (a) macro-microstructure; (b) microstructure at higher magnification ( the black equiaxed particles are α2 phases, the gray regions are O phases, and light regions are B2 phases).
290
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
cooling to room temperature after isothermally forging. In comparison with the as-forged alloy, more lamellar O phases precipitated and equiaxed α2 particles transformed into O or B2 matrix. The initial B2 grain size of the alloys produced by non-isothermal forging is well controlled and counted to be 37 μm–72 μm due to the pinning effect of the α2 particles; while the average B2 grain size of the alloys after isothermal forging at 1080 1C is about 320 μm. During heating and isothermal forging process at 1080 1C, the equiaxed α2 particles were gradually decreasing and go away completely. So the B2 grain size is greater in microstructures. Although the lamellar O phases could be obtained in isothermal forging processing at 1080 1C, the distribution of the lamellar O phases is inhomogeneous and the lamellar O phases are very fine. In order to obtain bimodal size lamellar O microstructures, the microstructures after isothermally forging at 1080 1C were solution- and age-treated in α2 þB2þ O and OþB2 phase regions. The microstructures and phases of Ti–22Al–25Nb alloy after water quenching from different solution temperatures are shown in Fig. 4(a) illustrates the microstructure of the isothermally forged Ti–22Al–25Nb alloy solution treated at 920 1C. It mainly contained lamellar O phases and B2 matrix. In comparison
with the as-isothermal forged alloy, the very fine-lamellar O phases were dissolved, the coarser-lamellar O phases formed and became wider and shorter, and some have already become globular during the heat-treatment. As the solution temperature increased to 960 1C (Fig. 4 (b)), more lamellar O phases dissolved into B2. The volume fraction of lamellar O decreased, and the shape became shorter and coarser. The amount of the O phase precipitates decreased obviously as the solution temperature increases to 980 1C which can be seen in Fig. 4 (c). This microstructure which is clearly different from the microstructures below 980 1C contains lamellar α2 /O and B2. Note that α2 phase formed as acicular in the lamellar O phase in 980 1C (this temperature is located in α2 þ B2þO phase field). This α2 phase has an orientation relationship to the parent O phase [12], ½001O ==½0001α2 and (110)o/ /(101̄ 0Þα2 [12]. The microstructure is very sensitive to the solution treatment temperature. When the alloys solution-treated in different phase region, the microstructures are different. The coarse-lamellar O phases were obtained in solution-treatment process. In order to realize the bimodal size lamellar O microstructures, aging treatments were performed on selected solution-treated samples. Fig. 5(a)–(e) illustrates the
Fig. 3. Phase and microstructure of the isothermally forged alloys at 1080 1C: (a) XRD image (b) SEM image (the black equiaxed particle is α2 phase, the gray is O phase and light regions are B2).
Fig. 4. Microstructures of solution-treated and water-quenched specimens. Solution-treatment temperatures were (a) 920 1C, (b) 960 1C and (c) 980 1C.
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
microstructures solution-treated at 940 1C and then age-treated between 760 1C and 840 1C for 12 h, followed by air cooling. All the aged microstructures were comprised of secondary acicular precipitates of the O phase within the B2 grains. The reason is believed to be the low-energy configuration for specific sets of planes in the O and B2 phases (i.e., OR) [13]. The size and volume fraction of the acicular precipitates of the O phase depend on the aging temperature. The largest precipitate size occurred at 840 1C and the finest size was displayed at 760 1C. It also can be seen from the figures that the amount of the secondary acicular O precipitates decreases as the age treatment temperature increases. From Fig. 5, noted that the more very fine acicular α2 phases are presented in lower age-treatment temperature (760 and 780 1C). As the temperature increased to 800– 840 1C, the amount of the fine acicular α2 phases is very rare and the width of α2 phases became thicker. Although the samples were agetreated in B2þO phase region for 12 h, the precipitation occurred with a lenticular morphology, and the primary lamellar O remained relatively unchanged. This is important for the design of the bimodal size lamellar O microstructures because the primary coarse-lamellar O phases and volume fraction can be controlled by selecting the appropriate solution temperatures and the secondary acicular precipitate size and volume fraction can be controlled by selecting the appropriate aging temperatures. According to the previous researches [8], the primary coarse-lamellar O make a greater contribution for the creep resistance, but exhibits poor strengthening effect compares to the fine secondary acicular O. Thus, to better adjust these microstructure parameters is believed to be important for obtaining the superior mechanical properties of the alloy.
291
and 10.2% for HT-760, 1116 MPa, 975 MPa, 10.5% and 11% for HT800, and 1066 MPa, 926 MPa, 12.5% and 12% for HT-840. Thus, the tensile strength of HT-760 was greater than that for HT-840; however it exhibits poorer process ability and lower elongation values. The various parameters which influence the yield strength are (1) the grain size of the B2 phase, (2) the strength of the individual phases as influenced by changes in their composition with heattreatment, (3) the volume fractions of α2/O and B2 phases, and (4) the lath size of O phase, (5) dislocation substructure [14]. No estimate has been made of the variation of the first factors in this study. The heat-treatment regimes of HT-760, HT-800, and HT-840 are the same except for the difference of age-treatment. The grain size of the B2 phase does not vary with age-treatment. Since the α2 phase volume fraction is extremely little for three microstructures and the tensile properties of Ti–22Al–25Nb alloy are significantly influenced by the microstructures, especially the volume fraction of B2 and O phases, and size of secondary lamellar O phase. From Table 2, it can be seen that increase in aging temperature leads to a decrease in strength and an increase in ductility. This variation trend of strength is caused by the noticeable coarsening besides the volume fraction reduction of secondary lamellar O phase with increasing aging temperatures [15]. The result is in agreement with literature [16], the yield strength is associated with the size and morphology of constituent phases. The yield strength can be expressed by the following rule of mixture:
σ a ¼ σ O f O þ σ 0B2 f B2 where σa is the yield strength of alloy; fo and fB2 designate the volume fraction of primary O phase and B2 phase, respectively; σo is the yield strength of O phase; σ B2' is a complex parameter that is determined by the size of B2 after thermo-mechanical processing and the volume fraction of O phase evolved in primary B2 matrix during heat treatment followed. For a given solutiontreatment temperature, the change of age-treatment temperature can influence the volume fraction of B2 and O phase, and the geometrical features of O phases.
3.2. Tensile properties The room temperature (RT) and 650 1C high temperature tensile properties, including Yield strength (YS), Ultimate tensile strength (UTS), Elongation (E), and Reduction area (RA), for selected microstructures are listed in Table 2. In most cases the results represent the average of at least three tests. The UTS, YS, E and RA of RT tensile properties were 1123 MPa, 997 MPa, 10%
Table 2 Tensile properties of the Ti–22Al–25Nb alloy (at%) at different heat-treatment temperatures. Testing temperature/1C
Sample
ST Temperature/1C
AT Temperature/1C
Ultimate tensile strength/MPa
Yield strength/MPa
Elongation/%
Reduction area/%
RT
HT-760 HT-800 HT-840
940 940 940
760 800 840
1123 1116 1066
997 975 926
10 10.5 12.5
10.2 11 12
650
HT-760 HT-800 HT-840
940 940 940
760 800 840
937 906 880
845 830 780
14 15 15.5
59 59 54
ST¼solution treatment and AT ¼ age treatment. Table 3 Heat treatment schedules and the corresponding microstructural parameters of the Ti–22Al–25Nb alloy. Vp, Llath, Wlath, Lacicular, Wacicular and d* are the volume fraction of each phase, the length of primary coarse-lamellar O, the width of primary coarse-lamellar O, the length of secondary fine acicular O, the width of secondary fine acicular O, and the grain size of B2 phase, respectively. ε is the average errors. Sample
α2 Vp ( 7 ε)
O Vp ( 7ε)
B2 Vp ( 7ε)
dn (μm) ( 7ε)
Llath /μm ( 7 ε)
Wlath /μm ( 7 ε)
Lacicular /μm ( 7ε)
Wacicular /μm ( 7 ε)
HT-760
1.85 (0.05)
72.98 (2.2)
25.17 (1.3)
217 (5.4)
4.04 (0.2)
0.28 (0.02)
0.91 (0.04)
0.09 (0.003)
HT-800
0.97 (0.04)
67.68 (1.8)
31.35 (1.3)
219 (6.1)
3.83 (0.4)
0.37 (0.03)
1.47 (0.07)
0.16 (0.004)
HT-840
0.31 (0.02)
63.62 (2.3)
36.07 (1.1)
220 (5.8)
2.15 (0.3)
0.42 (0.04)
2.13 (0.07)
0.23 (0.004)
292
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
From Table 3, it can be seen that the volume fraction of O phase, B2 phase and α2 phase are 72.98%, 25.17%, and 1.85% for HT-760; 67.68%, 31.35%, and 0.97% for HT-800; and 63.62%, 36.07%, and 0.31% for HT-840. It also can be seen that as the age temperature increased, the length and width of primary coarse-lamellar O became shorter and coarser, while the length and width of secondary acicular O became longer and coarser. The volume fraction of B2 for HT-760 is lowest than that for others and the volume fraction of O for HT-760 is highest, while the YS and UTS at RT of HT-760 outperformed HT-840. In the present work, tensile samples were solution treated at a constant temperature, so the volume fraction and the size of primary lamellar O are constant. The volume fraction of secondary acicular O phases increases as the aging decreases for the same solution temperatures. It means that the strength of alloy is mainly about precipitation strengthening of secondary acicular O-phases. The E values were 10% for the HT-760 RT tensile experiments, while the E values were 12.5% for the HT-840 RT tensile experiments. This difference of E values is mainly ascribed to the volume fraction of the O phase and B2 phase. O-phase platelets also provided a significant strengthening effect. Although the ordered O-phase is stronger than the B2-phase, it suffers from a lower number of available slip systems. Compared with the B2 structure, the O-phase suffers from a lower number of available slip systems [17]. Even though the
active slip systems between the O and BCC phases are compatible, the limited number of O-phase slip systems is considered to be responsible for the low E values in O-phase dominated microstructures. So the E values for the HT-840 RT outperformed the E values for the HT-760 RT. The UTS for HT-760 outperformed the UTS for HT-840, the higher O-phase volume fractions and finer O-phase laths were expected to be the probable reason for this behavior. However, Ophase dominated polycrystalline microstructures have proven to be brittle [3,16]. Thus, maintaining adequate volume fractions of the BCC phase is necessary for providing ductility. The fractographs of the room-temperature tensile specimens are shown in Fig. 6. It can be seen that the fracture mechanism of the material is a mixture of cleavage and faceted fracture. In addition, the specimens exhibited a flat fracture characteristic of brittle metals. The microstructures of alloys exhibited a combination of the fracture characteristics: both cleaved and faceted O grains and dimples within B2 regions. For HT-760, planar-slip traces were identified in the O phase, O/O cracking developed, while for HT-840, wavy slip was exhibited in B2 phase, slip compatibility between O and B2 grains was observed. The 650 1C tensile results are shown in Table 2. The tensile specimen for HT-760, which exhibited a YS¼ 845 MPa and a UTS¼937MPa, was significantly stronger than the tensile specimen for HT-840, which exhibited a tensile YS ¼780 MPa and a
Fig. 5. Comparison of a 940 1C/1 h solution-treated specimen and aged at different temperatures. The specimens were aged at (a) 760 1C, (b) 780 1C, (c) 800 1C, (d) 820 1C, and (e) 840 1C for 12 h followed by air cooling.
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
UTS¼ 880 MPa. For all temperatures, the YS decreased with increasing temperature for the alloys. The highest 650 1C YS value was exhibited in HT-760, greater O-phase concentrations tended to provide greater elevated-temperature strength. This suggests that the O phase provides greater elevated-temperature precipitation strengthening in these alloys.
293
coarser lamellar O and secondary acicular lamellar O phases became coarser. According to literature [2], the increased thickness of lamellar O-phase is beneficial to the improvement of the creep resistance of alloys. However, the facts are to the contrary. We
3.3. Creep property The creep behavior of the alloy for three heat treatment schedules has also been studied. Fig. 7 shows the creep curves of the three microstructures at 150 MPa and 650 1C for 100 h. Each plot possesses the typical primary or transient creep stage followed by a steady-state stage. No obvious tertiary creep stage was observed. The shape of creep curves is essentially the same, independently of the three microstructural variables. It can be seen that the creep rate of the primary creep stage for HT-840 is higher than that of HT-800 and HT-760. HT-760 has best creep resistance at 650 1C/150 MPa as compared with others. It mainly relates to the microstructures. From Table 3, the volume fraction of α2 phase is much less, the most volume fraction of α2 phase for HT-760 is only 1.85%, so we can probably ignore the impact of α2 phase on mechanical properties. For the microstructure of the HT-760 and the HT-840, as the age temperature increased, the width of primary
Fig. 7. The creep curves of the three heat-treatment schedules at 150 MPa and 650 1C for 100 h.
Fig. 6. Room-temperature fractographs for the samples: (a) macro-fractograph for samples HT-760; (b), (c) micro-fractograph for samples HT-760; (e) macro-fractograph for samples HT-840; (f), and (g) micro-fractograph for samples HT-840.
294
W. Wei et al. / Materials Science & Engineering A 618 (2014) 288–294
Fig. 8. TEM images of creep sample at 650 1C/150 MPa (a) HT-840 and (b) HT-760.
conjecture that except for the lamellar O phase, the volume fraction of B2 phase can also significantly influence the creep properties. For the three microstructures, the volume fraction of B2 for HT-840 is higher than that of HT-760, while the volume fraction of O for HT840 is lower than that of HT-760. The lamellar O phases have better creep resistance than B2 phases and the volume fraction of secondary acicular lamellar O phases for HT-760 is higher than that of HT-800 and HT-840, so HT-760 has the best creep resistance. In order to further study the effect of microstructure on creep behavior, the bright-field TEM images of the creep sample HT-760 and HT-840 have been present in Fig. 8. From Fig. 8, it can be seen that the creep resistance of coarser- and fine-lamellar O phase is different. The density of dislocations in the fine-lamellar O-phase seems to be very high, while the density of dislocations in the coarser-lamellar O-phase is low. It means the movement of dislocations in coarser-lamellar O-phase is not so active during creep deformation at 650 1C; so the creep resistance of coarserlamellar O-phase is higher than that of fine-lamellar O-phase. Fig. 8(a) shows the dislocations observed in crept sample HT-840. The microstructure is comprised of coarser- and fine-lamellar O phase, and B2 matrix. Because of more B2 phases present in this microstructure, the coarser- lamellar O was bent. From the microstructure for HT-760 (Fig. 8(b)), tangled or bowing dislocations and any kink or jog are not observed. Only few straight dislocations can be observed in the crept alloy. It means the movement of dislocations in this condition is not so active during creep deformation. No microstructural observations were recorded to suggest that grain boundary sliding contributes to the creep strains in this study. So, it was therefore suggested that the creep mechanism of HT-760 are dominated by a dislocation-controlled creep process, and this conclusion is supported by the observation of the dislocation densities. The current observations were insufficient to determine whether a glide or climb mechanism was rate controlling. 4. Conclusion In this study, microstructure evolution and tensile and creep properties of designing bimodal size lamellar O phases by thermomechanical processing including common forging, isothermal forging process and heat treatment for Ti–22Al–25Nb (at%) orthorhombic alloy were investigated. The effect of age-treatment temperature on microstructure and tensile and creep properties are also investigated. The microstructure of the isothermally forged alloys at 1080 1C contained the non-uniform distribution of lamellar O phases and B2 matrix. The lamellar O could be obtained when solution treated at OþB2 phase region and the secondary acicular O was emerged during the aging process. The volume fraction and thickness of coarse and acicular lamellar O could be well controlled by the solution and aging treatment. During aging treatments, the size and volume fraction of secondary acicular lamellar O are very
sensitive to temperature. The volume fraction decreased and the size increased with increasing of aging temperature. The size of secondary acicular lamellar O phase in α2 þ OþB2 phase region solution treatment plus aging condition is larger than that in OþB2 phase region solution treatment plus aging treatment. The size and volume fraction of secondary acicular lamellar O affect the mechanical properties of the alloy strongly. In terms of the RT tensile behavior, the YS and UTS at RT of HT-760 outperformed HT-840; however the E and RA of HT-760 have a lower value. The trends of 650 1C tensile properties were similar to those at RT; however, each of the samples exhibited lower strength and higher E and RA values than at RT. In terms of the creep behavior, HT-760 has best creep resistance at 650 1C/150 MPa as compared with HT-800 and HT-840. The creep properties of alloy were concerned with the microstructures. The volume fraction of B2 for HT-840 is higher than that of HT-760, while the volume fraction of O for HT-840 is lower than that of HT-760. The lamellar O phases have better creep resistance than B2 phases, so HT-760 has best creep resistance.
Acknowledgments This work was financially supported by Research Fund for the Doctoral Program of Higher Education of China with No. 201161 02110015, the New Century Excellent Talents in University with No. NCET-07-0696, and the National 973 Project of China with No. 2007CB613807. References [1] D. Banerjee, A.K. Gogia, T.K. Nandy, V.A. Joshi, Acta Metall. 36 (1988) 871. [2] A.K. Gogia, T.K. Nandy, D. Banerjee, T. Carisey, J.L. Strudel, J.M. Franchet, Intermetallics 6 (1998) 741. [3] C.J. Boehlert, D.B. Miracle, Metall. Mater. Trans. A 30 (1999) 2349–2367. [4] M. Hagiwara, T. Kitaura, Y. Ono, T. Yuri, T. Ogata, S. Emura, Mater. Trans. 53 (2012) 1138–1147. [5] Chen Xue, Weidong Zeng, B.i.n. Xu, Xiaobo Liang, Jianwei Zhang, Shiqiong Li, Intermetallics 29 (2012) 41–47. [6] L. Germann, D. Banerjee, J.Y. Guédou, J.-L. Strudel, Intermetallics 13 (2005) 920–924. [7] Wei Wang, Weidong Zeng, Chen Xue, Xiaobo Liang, Jianwei Zhang, Intermetallics 45 (2014) 29–37. [8] C.J. Cowen, C.J. Boehlert, Intermetallics 14 (2006) 412–422. [9] Jörg Kumpfert, Adv. Eng. Mater. 11 (2001) 851–864. [10] Y.W. Kim, D.M. Dimiduk, J. Manag. 43 (1991) 40–47. [11] Chen Xue, Weidong Zeng, Wei Wang, Xiaobo Liang, Jianwei Zhang, Mater. Sci. Eng. A 587 (2013) 54–60. [12] K. Muraleedharan, D. Banerjee, S. Banerjee, S. Lele, Philos. Mag. 71 (1995) 1011–1036. [13] C.J. Boehlert, B.S. Majumdar, V. Seetharaman, D.B. Miracle, Metall. Mater. Trans. A 30 (1999) 2305–2323. [14] A.K. Gogia, T.K. Nandy, K. Muraleedharan, D. Banerjee, Mater. Sci. Eng. A 159 (1992) 73–86. [15] C.L. Li, X.J. Mi, W.J. Ye, S.X. Hui, Y. Yu, W.Q. Wang, J. Alloy Compd. 550 (2013) 23–30. [16] C.J. Boehlert, D.B. Miracle, Metall. Mater. Trans. A 38 (2007) 26–34. [17] C.J. Boehlert, Metall. Mater. Trans. A 32 (2001) 1977–1988.