2.1 Effect of Mo element on microstructures of Mo-TiAl alloys
Fig.1 shows XRD patterns of the four Mo-TiAl alloys with different Mo contents. It can be seen that the four specimens are composed of γ-TiAl, α2-Ti3Al, and β phase. The diffraction peak at 2θ=38.6° represents the γ-(111) phase and has the highest intensity, indicating the largest volume fraction of the γ phase among these phases in the alloys. With increasing the Mo content, the diffraction peak intensity of β/B2-(110) phase (2θ=39.7°) is gradually increased, indicating that the content of β phase in Mo-TiAl alloys is increased significantly.
Fig.1 XRD patterns of different as-cast Mo-TiAl alloys
Fig.2 shows the as-cast microstructures of the four Mo-TiAl alloys. Due to the difference in atomic number, the γ, α2, and β phases show different contrast colors: the bright white is the β phase, the dark black is the γ phase, and the gray is α2 phase. The bright white β phase is mainly distributed at the lamellae boundary. The γ phase mainly exists in the form of γ/α2 lamellae, and some of them are distributed at the phase boundary. As for the α2 phase, it mainly appears in the form of γ/α2 lamellae, so it is difficult to solely observe.
Fig.2 BSE microstructures of different as-cast Mo-TiAl alloys: (a) TNM1, (b) TNM2, (c, e) TNM3, and (d, f) TNM4
It can be seen from Fig.2a that β and γ phases are scattered around the γ/α2 lamellae. With increasing the Mo content to 1.59at%, the content of β and γ phases around the lamellae is increased, and the increase in β phase content is particularly obvious (Fig.2b).
As the Mo content further increases, some of the γ/α2 lamellae in TNM3 and TNM4 alloys are transformed into the cluster structure, as shown in Fig.2c and 2d, respectively. Fig.3 shows TEM microstructure and selected area electron diffraction (SAED) pattern of the cluster. It can be seen that the cluster surrounding the γ/α2 lamellae in Mo-TiAl alloys consists of coexistent γ+α2+β phases, which is mainly caused by the chemical imbalance during the smelting process. The existence of the γ+α2+β coexistent phase refines the microstructure to a certain extent. Cha et al[18] found a coexistent phase, which is similar to that in this research, in the Mo-TiAl alloy after heat treatment at eutectoid temperature. The coexistent phases have the following coherent relationship: {110}β0<111>β0//(20)ω[0001]ω//(0001)α2[20]α2//{111}γ<110]γ. Thus, it can be inferred that the coexistent phase is formed by consuming α2 and γ phases in the γ/α2 lamellae.
Fig.3 TEM microstructure and SEAD pattern of coexistent γ+α2+β phase
Fig.4 shows the SEM microstructures of the four Mo-TiAl alloys after HIP. TNM1 and TNM2 specimens have relatively uniform microstructure, while the dark band-shaped regions can be observed in TNM3 and TNM4 alloys, indicating the existence of element segregation. Generally, there are three types of element segregation in TiAl alloys: β-type segre-gation, α-type segregation, and S-type segregation[19-21]. The β-type segregation mainly occurs at the periphery of γ/α2 lamellae, and it originates from the high-Nb area caused by the transformation from β to α phase. The α-type segregation mainly occurs inside the lamellae. When the content of β phase stabilization elements is high, the α→α2+γ+β reaction occurs, resulting in the precipitation of β phase particles inside the lamellae. The S-type segregation is formed by the segregation of Al elements in the dendrite during the solidification process, and it is often accompanied with uneven lamellar structure and lamellae coarsening. The dark areas in Fig.4c and 4d indicate the area with concentrated Al element, i.e., the S-type segregation occurs in TNM3 and TNM4 specimens. This S-type segregation exists in Mo-TiAl alloys after HIP treatment, and it will adversely affect the mechanical properties of Mo-TiAl alloys.
Fig.4 SEM microstructures of different Mo-TiAl alloys after HIP: (a) TNM1, (b) TNM2, (c) TNM3, and (d) TNM4
Fig.5 shows the magnified microstructures of Mo-TiAl alloys after HIP. It can be seen that the dark areas around the γ/α2 lamellae in these four Mo-TiAl alloys are enlarged, compared with those in the as-cast Mo-TiAl alloys, indicating that the content of γ phase is slightly increased. The β and γ phases appear as sharp strips around the γ/α2 lamellae in the as-cast alloys, whereas they are partially spheroidized in the form of block or sphere in the alloys after HIP. It can be seen from Fig.5c that the volume fraction of the γ+α2+β coexistent phase in TNM3 specimen significantly reduces. Fig.5d shows that the volume fraction of γ/α2 lamellar phase in TNM4 specimen decreases significantly, and the microstructure is mainly composed of equiaxed γ and β phases. Besides, a large amount of β phase precipitation appears in the γ/α2 lamellae.
Fig.5 Magnified microstructures of different Mo-TiAl alloys after HIP: (a) TNM1, (b) TNM2, (c) TNM3, and (d) TNM4
2.2 Enrichment behavior of β-phase stabilization elements in each phase
The γ, α2, and β phases in the initial TiAl alloys have different chemical composition. After the Mo addition, the chemical composition of these three phases also changes. β phase is easy to detect, since it is distributed at the lamellae boundary. The γ and α2 phases can only be distinguished in the magnified images, as shown in Fig.6.
Fig.6 SEM images for phase component analyses of different Mo-TiAl alloys: (a) TNM1, (b) TNM2, (c) TNM3, and (d) TNM4
The phase composition of Mo-TiAl alloys before and after HIP is shown in Table 2 and Table 3, respectively. It should be mentioned that due to the small content of B and Y elements, the measurement error is considerable. Thus, only the four main elements (Ti, Al, Nb, and Mo) are statistically analyzed.
Table 2
Chemical composition of γ, β, and α2 phases in different as-cast Mo-TiAl alloy specimens (at%)
Phase | Specimen | Ti | Al | Nb | Mo |
γ |
TNM1 |
48.34 |
46.66 |
3.75 |
0.94 |
TNM2 |
48.13 |
47.71 |
3.23 |
1.01 |
TNM3 |
46.81 |
48.14 |
3.58 |
1.25 |
TNM4 |
46.15 |
47.95 |
3.49 |
2.19 |
β |
TNM1 |
52.87 |
39.92 |
4.26 |
2.61 |
TNM2 |
55.39 |
37.61 |
4.02 |
2.66 |
TNM3 |
52.76 |
37.46 |
4.35 |
5.22 |
TNM4 |
50.41 |
38.45 |
4.18 |
6.77 |
α2 |
TNM1 |
52.78 |
42.63 |
3.55 |
1.04 |
TNM2 |
54.56 |
41.06 |
3.09 |
0.92 |
TNM3 |
55.93 |
38.75 |
3.69 |
1.63 |
TNM4 |
50.16 |
44.48 |
3.38 |
1.98 |
Table 3
Chemical composition of γ, β, and α2 phases in different Mo-TiAl alloy specimens after HIP (at%)
Phase | Specimen | Ti | Al | Nb | Mo |
γ |
TNM1 |
47.97 |
47.47 |
3.59 |
0.98 |
TNM2 |
48.06 |
47.47 |
3.43 |
0.75 |
TNM3 |
46.42 |
48.91 |
3.12 |
1.23 |
TNM4 |
44.92 |
49.15 |
3.59 |
1.98 |
β |
TNM1 |
54.57 |
37.94 |
4.39 |
2.84 |
TNM2 |
55.17 |
37.08 |
4.31 |
3.13 |
TNM3 |
51.93 |
38.06 |
4.01 |
5.67 |
TNM4 |
49.89 |
38.35 |
3.72 |
7.68 |
α2 |
TNM1 |
52.18 |
43.64 |
3.27 |
0.71 |
TNM2 |
53.52 |
42.29 |
3.20 |
0.90 |
TNM3 |
55.61 |
39.54 |
3.16 |
1.57 |
TNM4 |
49.79 |
44.93 |
3.18 |
1.97 |
It can be seen from Table 2 that the content of Ti and Al in the γ phase of the as-cast Mo-TiAl alloys is very close, and the ratio of Ti/Al is within 0.96~1.04. With increasing the Mo content, the ratio of Ti/Al in γ phase is gradually decreased. The ratio of Ti/Al in β phase of as-cast Mo-TiAl alloys is 1.31~1.47. The Nb content in the β phase is 4.02at%~4.35at%, which is higher than that in the γ phase (3.23at%~3.75at%) and in the α2 phase (3.09at%~3.69at%), indicating that Nb is enriched in the β phase. With increasing the Mo content, the Mo content in the β phase is increased obviously, which is about 3~4 times larger than that in γ phase and α2 phase. Therefore, it can be concluded that the Mo element mainly exists in the β phase.
In the α2 phase, the ratio of Ti/Al is 1.11~1.44. Besides, the content of Nb and Mo in the α2 phase is close to that in the γ phase. It should be noted that with increasing the Mo content, the enrichment of Nb element in the β phase is barely affected.
According to Table 3, it can be found that after HIP treatment, the Mo element content in the γ and α2 phase decreases, while that in the β phase further increases. This phenomenon implies that the Mo element is diffused from the γ and α2 phases to the β phase during HIP process. According to Table 2 and Table 3, with increasing the Mo content, the Ti content in the γ phase shows an overall downward trend, indicating that Mo atoms replace the Ti atoms in the γ phase matrix. Jiang et al[22] used the first-principles theory to calculate the occupancy of 3d, 4d, and 5d transition metal atoms in TiAl alloys, and found that the atoms, such as Zr and Hf, preferentially occupy the positions of Ti atoms. For Nb and Mo atoms, their occupancy status changes with their doping concentration and temperature. Huang et al[23] used the tight-binding energy band method to calculate the bonding strength of TiAl alloys after adding alloying elements, and predicted that most Mo atoms tend to replace Ti atoms, while a few Mo atoms replace Al atoms. Holec et al[24] performed the element doping calculations for binary TiAl alloys and found that the V, Cr, Nb, and Mo elements preferentially replace Ti atoms in γ and α2 phases, whereas they preferentially replace Al atoms in β phase. The B and C elements are located in the interstitial positions of Ti-rich octahedrons. These calculation results are consistent with the results in this research.
2.3 Nanoindentation hardness
In the nanoindentation test, most of the test points are on the γ/α2 lamellar phases, and the nanoindentation hardness is closely related to the size of the γ/α2 interlamellar space. It should be noted that the volume fractions of γ/α2 lamellae in these four Mo-TiAl alloys are different. In order to accurately describe the relationship between the nanoindentation hardness and the γ/α2 interlamellar spacing, it is necessary to exclude the difference in hardness caused by the microstructure changes. The average nanoindentation hardness of the TNM1, TNM2, TNM3, and TNM4 specimens is 5.67, 6.02, 5.81, and 5.56 GPa, respectively.
Fig.7 shows TEM microstructures of γ/α2 lamellae in the Mo-TiAl alloys, and the interlamellar space can be measured by the line-cutting method. The measured interlamellar spacing of TNM1, TNM2, TNM3, and TNM4 specimens is 417±24, 338±16, 379±21, and 423±29 nm, respectively. It is found that adding the Mo element within 1.59at% can refine the interlamellar spacing. However, when the Mo addition is larger than 1.59at%, the γ/α2 interlamellar spacing increases. Thus, the interlamellar spacing is negatively related to the nanoindentation hardness, i.e., the larger the interlamellar spacing, the smaller the nanoindentation hardness.
Fig.7 TEM microstructures of γ/α2 lamellae in different Mo-TiAl alloys after HIP: (a) TNM1, (b) TNM2, (c) TNM3, and (d) TNM4
Wang et al[25] studied the nanoindentation hardness of the directionally solidified 4722 alloy, and found that the nanoin-dentation hardness (Hnano) and the interlamellar spacing have the power function relationship, as follows:
where a and b are constants, and λ is the interlamellar spacing (nm). Substituting the nanoindentation hardness of TNM1, TNM2, TNM3, and TNM4 specimens and the measured interlamellar spacing into Eq.(1), the values of a and b can be obtained by line fitting method as 40.01 and -0.325 (R2=0.968), respectively. Thus, the relationship between nanoin-dentation hardness and interlamellar spacing can be expressed as follows:
In the nanoindentation test, the indenter was pressed into the Mo-TiAl matrix, and the plastic deformation occurs near pressing position of the indenter. The main plastic deformation mechanism is the dislocation slip. When the interlamellar spacing is short, the dislocations encounter more γ/α2 boundaries, increasing the hardness
2.4 Compression deformation behavior
Fig.8 shows the flow stress-strain curves of different Mo-TiAl alloys after HIP during hot compression deformation at 1200 ℃ with a strain rate of 0.001 s-1. The flow stresses of these four Mo-TiAl alloys all increase firstly and then slowly decrease. As the Mo content increases, the peak flow stress of the Mo-TiAl alloys gradually decreases. The change of the peak flow stress is related to the volume fraction variation of γ/α2 lamellae. The average hardness of γ/α2 lamellae is higher than that of the equiaxed γ and α2 phases. The more the γ/α2 lamellae in the alloy, the higher the strength and the more obvious the work hardening phenomenon. Correspondingly, the peak stress rises.
Fig.8 Flow stress-strain curves of different Mo-TiAl alloys after HIP
Table 4 shows the surface cracking states of the four Mo-TiAl alloys after hot compression deformation at 1200 °C with different strain rates. It can be seen that the Mo-TiAl alloys have a tendency to crack during compression deformation at high strain rate (1 s-1). As the strain rate decreases, the surface tends to be complete. For TNM3 and TNM4 specimens, when the strain rate increases to 0.01 and 0.001 s-1, respectively, the surface cracks, indicating that their hot working deformation ability is inferior. According to Table 4, the hot workability of Mo-TiAl alloys is ranked from the best to the worst: TNM2>TNM1>TNM3>TNM4.
Table 4
Surface cracking of different Mo-TiAl alloys after HIP during hot compression deformation at 1200 °C with different strain rates
Strain rate/s-1 | 1 | 0.1 | 0.01 | 0.001 |
TNM1 |
× |
√ |
√ |
√ |
TNM2 |
√ |
√ |
√ |
√ |
TNM3 |
× |
× |
× |
√ |
TNM4 |
× |
× |
× |
× |
Note: × means that the specimen surface is cracked; √ means that the specimen surface is complete without obvious cracks
Fig.9 shows the microstructures of different as-HIPed Mo-TiAl alloys after hot compression deformation. Under the hot compression deformation condition of 1200 ℃/0.01 s-1, the lamellae in TNM1 and TNM2 specimens are transformed into the equiaxed α phase, which is relatively uniform. In TNM2 specimen, the residual β phase perpendicular to the hot compression direction can also be observed (Fig.9a and 9b). In TNM3 and TNM4 specimens, the microstructure is mainly composed of equiaxed γ and β phases, and the black mush zones appear after compression deformation. It can be seen that these mush zones consist of some residual γ/α2 lamellae and equiaxed γ phase. Around these mush zones, a large number of microcracks appear, and the direction of crack initiation and propagation is consistent with the orientation of the γ/α2 lamellae.
Fig.9 Microstructures of different as-HIPed Mo-TiAl alloys after hot compression deformation: (a) TNM1, (b) TNM2, (c) TNM3, and (d) TNM4
Fig.10 shows the SEM microstructure and energy disperse spectroscope (EDS) line scanning results of the mush zone in TNM3 specimen. The mush zone is rich in Al, and has a small amount of Ti, Nb, and Mo (Fig.10b), which is closely related to the S-type segregation in the Mo-TiAl alloys. The existence of microcracks near the mush area explains the sawtooth fluctuations in the flow stress of TNM3 and TNM4 specimens (Fig.8) [26]. This phenomenon also corresponds to the inferior hot workability of TNM3 and TNM4 specimens. For TNM1 specimen, its hot workability is not as good as that of TNM2 specimen, mainly because TNM2 specimen contains more β phases. The β phase has a body-centered cubic structure, so the dislocations are easy to initiate, thus playing a coordi-nating role in the deformation process. Meanwhile, according to the hot compression performance of TNM3 and TNM4 specimens, the more β phases in alloy do not necessarily lead to the better plasticity of alloy. On the one hand, during high-temperature deformation, the β phase needs to be introduced, but it is also necessary to fully consider the uniformity of the microstructure and chemical composition to avoid composi-tion segregation caused by the excessive addition of β phase stabilization elements. On the other hand, too much β phase will be transformed into the B2 phase at room temperature, resulting in inferior plasticity and creep properties at room temperature. If the β/B2 phase cannot be controlled by effective heat treatment, it is difficult for TiAl alloys to achieve ideal mechanical properties at room temperature.
Fig.10 SEM microstructure (a) and EDS line scanning along marked greenline (b) in TNM3 specimen (for display convenience, the intensity of EDS spectra of Nb and Mo elements is increased by 5 times than the original intensity)
1) The TiAl alloys containing Mo element of 1.02at%, 1.59at%, 2.11at%, 3.94at% are named as TNM1, TNM2, TNM3, and TNM4 specimens, respectively, which are composed of γ, α2, and β phases. As the content of Mo element increases, the content of the γ phase is gradually decreased, while that of the β phase is increased significantly. After the hot isostatic pressing (HIP) treatment, the γ phase around the lamellae increases, the TNM1 and TNM2 specimens have uniform microstructures, while the S-type segregation exists in the TNM3 and TNM4 specimens.
2) The atomic ratios of Ti/Al in the γ, α2, and β phases are 0.96~1.04, 1.11~1.44, and 1.31~1.47, respectively. The added Mo element mainly exists in the β phase, and the content of Mo element in the β phase is increased with increasing the Mo addition. After HIP, the Mo content in the γ and α2 phases is decreased slightly, and that in the β phase is increased, indicating that Mo element is diffused from the γ and α2 phases to the β phase.
3) With the addition of Mo element, the nanoindentation hardness of the Mo-TiAl alloys is firstly increased and then decreased, reaching the maximum hardness with the Mo content of 1.59at%. The relationship between nanoindentation hardness and interlamellar spacing (λ) of Mo-TiAl alloys is .
4) The addition of the Mo element decreases the flow stress of Mo-TiAl alloys. During the hot compression at a high strain rate, the microcracks appear on the surface of TNM1, TNM3, and TNM4 specimens. The hot deformation ability of these four specimens from excellent to poor is arranged as follows: TNM2>TNM1>TNM3>TNM4. The poor hot workability of TNM3 and TNM4 specimens is due to the element segregation and uneven microstructure in alloys.
References
1 Li H, Yang C, Sun L X et al. Materials Letters[J], 2017, 187: 4 [Baidu Scholar]
2 Sallot P, Monchoux J P, Joulié S et al. Intermetallics[J], 2020, 119: 106 729 [Baidu Scholar]
3 Tian S W, Jiang H T, Guo W Q et al. Intermetallics[J], 2019, 112: 106 521 [Baidu Scholar]
4 Li Haiyan, Kou Peipei, Li Longlong. Rare Metal Materials and Engineering[J], 2020, 49(5): 1593 [Baidu Scholar]
5 Li Haiyan, Qiao Haiyang, Feng Ruicheng et al. Rare Metal Materials and Engineering[J], 2020, 49(6): 1931 [Baidu Scholar]
6 Song L, Appel F, Wang L et al. Acta Materialia[J], 2020, 186: 575 [Baidu Scholar]
7 Chlupová A, Heczko M, Obrtlík K et al. Materials Science and Engineering A[J], 2020, 786: 139 427 [Baidu Scholar]
8 Gu X, Cao F Y, Liu N et al. Journal of Alloys and Compounds[J], 2020, 819: 153 264 [Baidu Scholar]
9 Yadav M K, Siddiquee A N, Khan Z A. Metals and Materials International[J], 2021, 27(7): 2378 [Baidu Scholar]
10 Jiang H T, Tian S W, Guo W Q et al. Materials Science and Engineering A[J], 2018, 719: 104 [Baidu Scholar]
11 Imayev R M, Imayev V M, Oehring M et al. Intermetallics[J], 2007, 15(4): 451 [Baidu Scholar]
12 Banumathy S, Neelam N S, Chandravanshi V et al. Materials Today: Proceedings[J], 2018, 5(2): 5514 [Baidu Scholar]
13 Han J C, Xiao S L, Tian J et al. Materials Characterization[J], 2015, 106: 112 [Baidu Scholar]
14 Han J C, Liu Z D, Jia Y et al. Vacuum[J], 2020, 174: 109 210 [Baidu Scholar]
15 Chen Y Y, Li B H, Kong F T. Journal of Alloys and Compounds[J], 2008, 457(1-2): 265 [Baidu Scholar]
16 Xu W C, Zhang H, Shan D B. Journal of Zhejiang University-Science A[J], 2010, 11(10): 738 [Baidu Scholar]
17 Clemens H, Mayer S. Advanced Engineering Materials[J], 2013, 15(4): 191 [Baidu Scholar]
18 Cha L, Clemens H, Dehm G. International Journal of Materials Research[J], 2011, 102(6): 703 [Baidu Scholar]
20 Liu G H, Wang Z D, Fu T L et al. Journal of Alloys and Compounds[J], 2015, 650: 45 [Baidu Scholar]
21 Dong S L, Liu T, Li Y J et al. Vacuum[J], 2019, 159: 391 [Baidu Scholar]
22 Jiang C. Acta Materialia[J], 2008, 56(20): 6224 [Baidu Scholar]
23 Huang Zunixng, Wang Xiuli, Zhou Lixin et al. Chinese Journal of Structural Chemistry[J], 2002, 21(2): 218 (in Chinese) [Baidu Scholar]
24 Holec D, Reddy R K, Klein T et al. Journal of Applied Physics[J], 2016, 119(20): 205 104 [Baidu Scholar]
25 Wang Qiang. Thesis for Doctorate[D]. Harbin: Harbin Institute of Technology[D], 2018 (in Chinese) [Baidu Scholar]
26 Cao Hui, Rui Zhiyuan, Chen Wenke et al. Molecular Simulation[J], 2018, 44(18): 1489 [Baidu Scholar]