Eutectic evolution of directionally solidified Nb-Si based ultrahigh temperature alloys

Eutectic evolution of directionally solidified Nb-Si based ultrahigh temperature alloys

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279 Contents lists available at ScienceDirect International Journal of Ref...

4MB Sizes 2 Downloads 52 Views

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

Contents lists available at ScienceDirect

International Journal of Refractory Metals & Hard Materials journal homepage: www.elsevier.com/locate/IJRMHM

Eutectic evolution of directionally solidified Nb-Si based ultrahigh temperature alloys ⁎

Na Wang, Lina Jia , Bin Kong, Yueling Guo, Huarui Zhang, Hu Zhang

T



School of Materials Science and Engineering, Beihang University, Beijing 100191, China

A R T I C L E I N F O

A B S T R A C T

Keywords: Nb-Si alloy Transition metal silicide Directional solidification Segregation Microstructure transformation

The purpose of this work is to reveal the eutectic evolution in near-eutectic Nb-Si based alloys. The Nb–xTi–ySi–4Cr–2Al–2Hf (22Ti-15Si; 24Ti-15Si; 22Ti-16.5Si, at.%) alloys were processed by directional solidification with the withdrawal rates of 400 and 500 μm/s. Three typical eutectics were observed: quasi-regular eutectics I, quasi-regular eutectics II and eutectic dendrites colonies. The decrease in the content of Ti and the increase in the withdrawal rate increased the growth rate, and led to the transformation from quasi-regular eutectics I to quasi-regular eutectics II. The increase in the content of Si and the decrease in the withdrawal rate reduced the growth rate, led to larger degree of segregation, and eventually promoted the generation of eutectic dendrites colonies. The sizes of quasi-regular eutectics I and II decreased with the increase in the withdrawal rate and increased with the higher contents of Ti and Si. For the quasi-regular eutectics I and II, the volume fractions of γ-(Nb, Ti)5Si3 phases increased with increasing the content of Ti. Compared with the quasi-regular eutectics I, more γ-(Nb, Ti)5Si3 phases existed in the quasi-regular eutectics II, due to the growth pattern transition of the γ(Nb, Ti)5Si3 phases.

1. Introduction With higher melting points (˃1750 °C), relatively lower densities (6.6–7.2 g/cm3) and more attractive stiffness and high-temperature strength than nickel (Ni) alloys, Nb-Si based alloys show great promise as the next generation turbine airfoil materials at the temperature of up to 1300 °C [1,2,3]. According to the Nb-Si binary phase diagram [4,5], a (Nb)/Nb5Si3 two-phase microstructure generates through a eutectic reaction (L → (Nb) + Nb3Si) at 2193 K, and a eutectoid reaction (Nb3Si → (Nb) + Nb5Si3) at 2043 K. The brittle intermetallic Nb5Si3 phases provide high-temperature strength, while the bcc Nb solid-solution phase (Nbss) can improve the room temperature toughness of alloy matrix [6,7]. It has been one of the focuses of recent studies [8,9,10,11,12] that enhancing the low room temperature toughness of the Nb-Si based alloys by alloying, optimizing alloy composition and processing technologies. The aim of these methods is to optimize the microstructure and volume fraction of constituent phases, the determining factors for the room temperature fracture toughness. The high volume fraction of Nbss phases in the hypoeutectic Nb-Si based alloys, containing primary Nbss and eutectic Nbss/Nb5Si3, has an advantage in improving room temperature toughness [9,13]. However, it would impose negative influence on high-temperature strength [9,14]. On the contrary, the high volume fraction of Nb5Si3 phases in



hypereutectic Nb-Si based alloys is beneficial to improve high-temperature strength, but it sacrifices the room temperature toughness [13,14]. Therefore, the microstructural control of eutectic Nb-Si based alloys is an option for improving room temperature toughness and without the sacrifice of excellent high-temperature strength [14,15,16,17]. The various alloy compositions and processing technologies promote the generation of various morphologies of eutectic Nbss/Nb5Si3. In previous reports [13,15,17,18], there are mainly four types of eutectics observed in near-eutectic Nb-Si based alloys: dispersion eutectics, lamella eutectics, maze-like eutectics and the eutectics with columnar Nb5Si3. The dispersion eutectics have a negative impact on room temperature toughness [13]. The lamella eutectics [15,17], the maze-like eutectics [13] and the eutectics with columnar Nb5Si3 [18] are contributed to improve the room temperature toughness. The morphologies of eutectic Nbss/Nb5Si3 have a great influence on the room temperature toughness. However, only a few studies have been focused on the characterization of eutectic morphologies, eutectic evolution and eutectic microstructure control of Nb-Si based alloys [15,19,20,21]. Consequently, the mechanism of eutectic evolution and formation is unclear and it is difficult to improve toughness by microstructure control. The purpose of this work is to obtain different Nbss/Nb5Si3 eutectic

Corresponding authors. E-mail addresses: [email protected] (L. Jia), [email protected] (H. Zhang).

https://doi.org/10.1016/j.ijrmhm.2017.11.001 Received 16 May 2017; Received in revised form 5 September 2017; Accepted 2 November 2017 Available online 05 November 2017 0263-4368/ © 2017 Elsevier Ltd. All rights reserved.

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

and DS3 samples. The lightly contrasted regions are confirmed to correspond to (Nb, Ti)ss phases. The grayly and darkly contrasted regions correspond to (Nb, Ti)5Si3 phases. From the longitudinal sections of quasi-regular eutectic I (Fig. 2(a) and (b)), the (Nb, Ti)ss phases grow freely and display skewed and bifurcate morphologies. In addition, they reveal dendritic growth in the center and lamellar morphologies in the exterior margin. The (Nb, Ti)5Si3 phases grow in the residual space and display block or short rod morphologies. From the transverse sections of quasi-regular eutectic I (Fig. 2(c) and (d)), it shows bud-like morphologies and a small amount of (Nb, Ti)5Si3 fragments are distributed in the center of the bud-like eutectic. The sizes of bud-like eutectics in DS3 samples are generally larger than those in DS1 samples. The average diameters of bud-like eutectics in DS1 and DS3 samples (15.589 μm and 18.339 μm, respectively) are shown in Table 1. Fig. 3 shows the morphologies of quasi-regular eutectic II of the DS2, DS4 and DS6 samples. From the longitudinal sections of quasiregular eutectic II (Fig. 3(a), (b) and (c)), the (Nb, Ti)ss phases and the (Nb, Ti)5Si3 phases grow in parallel along the growth direction. The (Nb, Ti)ss phases reveal dendritic growth in the center and lamellar morphologies in the exterior margin, particularly in the DS4 and DS6 samples (Fig. 3(b) and (c)). From the transverse sections of the DS4 and DS6, the quasi-regular eutectic II shows bud-like morphologies and the bean-shape (Nb, Ti)ss phases are distributed in the center, as shown in Fig. 3(e) and (f). From the transverse sections of DS2, the quasi-regular eutectics II shows the same morphologies as the quasi-regular eutectics I (Fig. 2(c) and (d)), which is due to the poor development of the (Nb, Ti)ss phases dendritic growth. As shown in Table1, the average diameters of bud-like eutectics in DS4 and DS6 samples (14.423 μm and 14.879 μm, respectively) are larger than that in DS2 sample (12.711 μm). Fig. 4 shows the morphologies of eutectic dendrites colonies of DS5. From the longitudinal sections, the quasi-regular eutectics I grow among eutectic dendrites colonies (Fig. 4(a)). The (Nb,Ti)ss phases of the eutectic dendrites colonies have dendritic structures and pass through the (Nb, Ti)5Si3 phases as marked by white arrows in Fig. 4(b). The (Nb,Ti)ss phases dendrites are coarser in the exterior margin. From the transverse sections, eutectic dendrites colonies show flower-like morphologies. The (Nb, Ti)ss phases and (Nb, Ti)5Si3 phases are distributed in a lamellar structure, growing radially in a coupled manner in the center of the flower-like microstructures. The average diameter of flower-like eutectics in DS5 sample is considerably large, about 43.092 μm. Compared with DS1 and DS3 samples, the average diameter of bud-like quasi-regular eutectics I in DS5 sample is much larger (22.808 μm). The average diameters of bud-like eutectics in DS1-DS6 samples and flower-like eutectics in DS5 sample are shown in Table 1. The average diameter of the quasi-regular eutectics II is smaller than the quasiregular eutectics I. It is due to the increase in the growth rates with a larger withdrawal rate [23,24]. As mentioned above, the average diameters of the bud-like quasi-regular eutectics I and II increase with the increase in the contents of Ti and Si. Higher contents of Ti and Si lead to the decrease in the melting point of the near-eutectic Nb-Si based alloy based on the Nb-Ti-Si phase diagram [25], and therefore the degree of undercooling decreases. As a result, it reduces the solidification rates and promotes the generation of larger bud-like eutectics. As shown in Figs. 2, 3 and 4, the (Nb, Ti)5Si3 phases includes gray (Nb, Ti)5Si3 phases and black (Nb, Ti)5Si3 phases. The black (Nb, Ti)5Si3 phases are frequently accompanied by a few of cracks perpendicular to the solidification direction as marked by red arrow in the insert of Fig. 2(a). They possess higher Ti, Hf, Cr and Al contents than the gray (Nb, Ti)5Si3 phases, especially Ti content (about 25.80 and 14.83 at.%, respectively), as determined by WDS. As revealed by the XRD patterns (Fig. 1), the (Nb, Ti)5Si3 phases include mainly α-(Nb, Ti)5Si3 phases and γ-(Nb, Ti)5Si3 phases. They are further confirmed by TEM. Fig. 5 shows the TEM bright-field image and selected area electron diffraction (SAD) patterns, which are along [0001] zone axis of the γ-

morphologies in near-eutectic Nb–xTi–ySi–4Cr–2Al–2Hf (22Ti-15Si; 24Ti-15Si; 22Ti-16.5Si, at.%) alloys and, furthermore, to reveal the eutectic evolution with different withdrawal rates and different contents of Ti and Si. 2. Experimental The master alloy ingots, with different nominal compositions of NbxTi-ySi-4Cr-2Al-2Hf (22Ti-15Si; 24Ti-15Si; 22Ti-16.5Si, at.%), were prepared by vacuum non-consumable arc-melting. The master alloy rods with about 13 mm in diameter were cut from the ingot by electrodischarge machining (EDM) and then removed the oxidized surface using lathe. They were cleaned with alcohol and were assembled with a self-contributed Y2O3 crucible. The crucible was nominally 18 mm in outside diameter, 15 mm in inside diameter and 220 mm in length. The directional solidification (DS) experiments were carried out in liquid metal cooling (LMC) furnace as described in our previous work [22].The furnace chamber was heated once the vacuum level was 5 × 10− 3 Pa. Then the high purity (99.99 wt%) argon was backfilled into the furnace chamber to minimize contamination. The melt superheat temperature was set at 1980 °C. When the temperature reached the preset value and the melt were held for 20 min at this temperature, the DS experiments were performed with the withdrawal rates of 400 and 500 μm/s respectively and subsequently quenched into a liquid Ga-InSn bath. The withdrawal distance was set as 195 mm. The DS samples with different withdrawal rates and nominal contents of Ti and Si were marked as DS1 (22Ti-15Si, 400 μm/s), DS2 (22Ti-15Si, 500 μm/s), DS3 (24Ti-15Si, 400 μm/s), DS4 (24Ti-15Si, 500 μm/s), DS5 (22Ti-16.5Si, 400 μm/s) and DS6 (22Ti-16.5Si, 500 μm/s). The phases were identified by X-ray diffraction (XRD, D/ max-2500, Cu Kα). Microstructure and element content distribution were investigated using electron probe micro-analyzers (EPMA; Model JXA-8230 and JXA-8100) with wave dispersive spectroscope (WDS) and energy dispersive spectroscope (EDS). The microstructure was further studied using transmission electron microscope (TEM; JEM2100F) with an EDS detector. The average diameters of different eutectics and the volume fractions of (Nb,Ti)5Si3 phases were measured using Image-Pro Plus 6 software. For each sample condition, five BSE images at the same magnification (500 ×) were used. 3. Results and discussions Three typical eutectics: quasi-regular eutectics I, quasi-regular eutectics II and eutectic dendrites colonies are observed in near-eutectic Nb-Si based alloys. Fig. 1 shows the XRD patterns of the DS1-DS6 samples. The phases are mainly Nbss, α-Nb5Si3 phases (body centered tetragonal, D8l structure) and γ-Nb5Si3 phases (hexagonal, D88 structure). Fig. 2 shows the morphologies of quasi-regular eutectic I of the DS1

Fig. 1. XRD patterns of the directional solidification Nb-Si based alloys.

274

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

Fig. 2. BSE images of quasi-regular eutectics I on both longitudinal ((a) and (b)) and transverse ((c) and (d)) sections of DS1 and DS3 samples. (a) and (c) is the BSE images of DS1; (b) and (d) is the BSE images of DS3. The inserts of Fig. 2(a), (c) and (d) are the enlarged morphologies.

(b)

(a)

-(Nb, Ti)5Si3

-(Nb, Ti)5Si3 (Nb, Ti)ss

(c)

(d)

Table 1 Average diameters of bud-like quasi-regular eutectics I and II and flower-like eutectic dendrites colonies. Eutectics

22Ti-15Si

Quasi-regular eutectics I Quasi-regular eutectics II Eutectic dendrites colonies

DS1 DS2 –

24Ti-15Si 15.589 ± 2.217 μm 12.711 ± 2.119 μm –

DS3 DS4 –

22Ti-16.5Si 18.339 ± 2.383 μm 14.423 ± 2.683 μm –

DS5 DS6 DS5

22.808 ± 2.677 μm 14.876 ± 2.400 μm 43.092 ± 6.431 μm

body centered tetragonal crystal structure of α-(Nb,Ti)5Si3 phases. Therefore, cracks are not observed within α-(Nb,Ti)5Si3 phases.

(Nb, Ti)5Si3 phase (PDF#65-3599, hexagonal structure) and [001] zone axis of the α-(Nb, Ti)5Si3 phase (PDF#65-2781, body centered tetragonal structure) respectively. The contents of Ti in γ-(Nb, Ti)5Si3 phase and α-(Nb, Ti)5Si3 phase (about 23.35 and 13.25 at.%, respectively), determined by TEM-EDS, are similar to those of black (Nb, Ti)5Si3 phase and gray (Nb, Ti)5Si3 phase determined by WDS, respectively. Therefore, the black (Nb, Ti)5Si3 phases are γ-(Nb, Ti)5Si3 phases and the gray (Nb, Ti)5Si3 phases are α-(Nb, Ti)5Si3. According to the works of Knittel et al. [26], the γ-Nb5Si3 phases stabilized for a sum (at.% Hf + Ti) above 28 at.%, whereas the αNb5Si3 phases stabilized for a sum (at.% Hf + Ti) below 21–22 at.%. The γ-Nb5Si3 silicide is stabilized by high Ti + Hf content [26]. Our present works are consistent with their works. The sum (at.% Hf + Ti) of γ-Nb5Si3 phases is about 29.83 at.% (4.03 + 25.80 at.%) and the sum (at.% Hf + Ti) of α-Nb5Si3 phases is about 17.04 at.% (2.21 + 14.83 at.%) in our present works. As proposed by earlier papers [10,27], the γ-(Nb, Ti)5Si3 phases are closely related to Ti segregation at the solid–liquid interface front. When the amount of Ti segregation reaches a certain value, the γ-(Nb, Ti)5Si3 phases can be formed, which are isomorphous with Ti5Si3 phases [10,28,29]. As a result of the strong thermal expansion anisotropy of γ-(Nb, Ti)5Si3 phases and the remarkable difference in the thermal expansion coefficients between γ-(Nb,Ti)5Si3 and (Nb, Ti)ss, the thermal stress is considerably large [30,31]. Moreover, there are less slip systems in the hexagonal crystal structure of γ-(Nb,Ti)5Si3 phases [32]. These are key reasons for the formation of cracks within γ-(Nb, Ti)5Si3 phases. Compared with γ-(Nb, Ti)5Si3 phases, the thermal expansion anisotropy of α-(Nb, Ti)5Si3 phases is weak. The match degree of thermal expansion coefficients between Nbss and α-(Nb, Ti)5Si3 is higher. Accordingly, the thermal stress is small. In addition, there are more slip systems in the

3.1. Generation of quasi-regular eutectics I As mentioned above, for the quasi-regular eutectics I, the (Nb, Ti)ss phases grow with freedom, presenting skewed and bifurcate morphologies, and the (Nb, Ti)5Si3 phases grow in the residual space and reveal block or short rod morphologies. It indicates that the growth rate of (Nb, Ti)ss phases is larger compared with that of (Nb, Ti)5Si3 phases. Nbss is a non-faceted phase and the entropy of fusion is 9.6 Jmol− 1 K− 1 [33,34], and Nb5Si3 is a faceted phase and the entropy of fusion is 14.13–23.27 Jmol− 1 K− 1 [34,35]. Note that the Nb5Si3 phases possess a greater entropy of fusion and their growth needs larger undercooling [34]. Therefore, the growth of Nb5Si3 phases is restricted and they distribute in the residual space with block or short rod morphologies. 3.2. Generation of quasi-regular eutectics II The (Nb, Ti)ss phases and (Nb, Ti)5Si3 phases of the quasi-regular eutectics II grow in parallel along the growth direction. It indicate the growth rates of (Nb, Ti)5Si3 phases and (Nb, Ti)ss phases are commensurate. Compared with DS1 and DS3 samples, the withdrawal rates of DS2 and DS4 samples are larger. Accordingly, the growth rates of (Nb, Ti)5Si3 phases and (Nb, Ti)ss phases increase [23,24]. The increase in the growth rate of (Nb, Ti)5Si3 phases contributes to the transition from faceted growth to continuous growth, where the growth rate tends to be proportional to the undercooling [36,37,38]. Consequently, the growth rate of (Nb, Ti)5Si3 phases is nearly the same as that of (Nb, Ti) 275

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 3. BSE images of quasi-regular eutectics II on both longitudinal ((a), (b) and (c)) and transverse ((d), (e) and (f)) sections of DS2, DS4 and DS6 samples. (a) and (d) is the BSE images of DS2; (b) and (e) is the BSE images of DS4; (c) and (f) is the BSE images of DS6. The inserts of Fig. 3(d), (e) and (f) are the enlarged morphologies.

the solidification front shows a dendritic morphology and the eutectic colonies are developed behind it as a result of the constitutional and curvature undercooling [42,43,44,45]. Compared with DS5 sample, the withdrawal rate of DS6 sample is larger and its growth rate is higher. Therefore, the degree of segregation for the low melting points elements decreases. Consequently, the eutectic dendrites colonies disappeared and quasi-regular eutectics II are observed in DS6 sample.

ss phases. 3.3. Generation of eutectic dendrites colonies With the increase of Si content (16.5 at.%), the eutectic dendrites colonies are observed in DS5 sample. Compared with DS1 sample, the increase in the content of Si in DS5 sample leads to the decrease in the melting point [25], which decreases the undercooling and slows down the growth rate of crystals. For solute diffusion, two factors need to be considered: the diffusion time (ΔT/GV, where ΔT denotes undercooling degree, G denotes the thermal gradient, V denotes withdrawal rate) and the diffusion distance (approximately equal to λ1/2, where λ1 denotes the primary dendrite arm spacing), as previously mentioned by Liu et al. [39,40,41]. The two factors are related to the growth rate, which has a critical value. Below the critical growth rate, the increase of diffusion time with the decrease in the growth rate reduces the degree of segregation. Over the critical growth rate, the decrease in the growth rate leads to the increase both in the diffusion distance (λ1/2) and the degree of segregation. In our present work, the withdrawal rates (400 and 500 μm/s) are considerably large and the growth rate should be over the critical value. Therefore, as a result of the lower growth rate in DS5 sample compared with DS1 sample, the low melting point elements such as Ti and Cr are more likely to concentrate in the solidification front. It causes a negative temperature gradient and a considerable undercooling in front of the crystal, which destabilizes a planar interface of the eutectic growth front in favor of a dendritic one. As a result,

3.4. Dendritic growth of (Nb, Ti)ss phases In addition, the (Nb, Ti)ss phases of the quasi-regular eutectics I and II reveals dendritic growth in the center and lamellar morphologies in the exterior margin from longitudinal sections (Fig. 2(a), Fig. 2(b), Fig. 3(a), Fig. 3(b) and Fig. 3(c)). Compared with DS1 and DS2, the dendritic growth is more obvious in DS3, DS4 and DS6 samples. By analogy with eutectic dendrites colonies in DS5 samples, the segregation of low-melting elements, particularly Ti and Cr, is one of the main causes for the dendritic growth of the (Nb, Ti)ss phases of the quasiregular eutectics I and II. The content of Ti in DS3 sample is higher than that in DS1 sample. The content of Ti in DS4 sample and the content of Si in DS6 sample are higher than that in DS2 sample. They lead to the decrease in the melting point based on Nb-Ti-Si phase diagram [25] and the decrease in the degree of undercooling. The growth rates slow down and the degrees of segregation in DS3, DS4 and DS6 samples are larger, which promote the dendritic growth of (Nb, Ti)ss phases. In present works, the Nb-xTi-ySi-4Cr-2Al-2Hf (24Ti-15Si; 22Ti-16.5Si, at.%) alloys 276

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

Fig. 4. BSE images of eutectic dendrites colonies on both longitudinal (a, b) and transverse (c) sections of DS5 sample.

(a)

(c)

(b)

eutectic dendrites colonies

(a)

(b)

(c)

(d)

Fig. 5. TEM bright-field images ((a) and (c)) and selected area electron diffraction (SAD) patterns ((b) and (d)) of the α-(Nb, Ti)5Si3 phases ((a) and (b)) and the γ-(Nb, Ti)5Si3 phases ((c) and (d)) of the quasi-regular eutectics I in DS1 sample.

277

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

(a)

Fig. 6. BSE images of quasi-regular eutectics II on longitudinal sections of Nb-XTi-YSi-4Cr-2Al-2Hf (24Ti-15Si; 22Ti-16.5Si, at. %) alloys with 600 μm/s withdrawal rates. (a) 24Ti-15Si; (b) 22Ti-16.5Si.

(b)

eutectic points. For the quasi-regular eutectics I in DS1 and DS3 samples and the quasi-regular eutectics II in DS2 and DS4 samples, the volume fractions of γ-(Nb, Ti)5Si3 phases increase with the increase in the content of Ti. As mentioned above, the higher content of Ti leads to the decrease in the growth rate and the increase in the degrees of segregation. The inherent higher content of Ti and the larger degree of segregation contributed in the generation of γ-(Nb, Ti)5Si3 phases in DS3 and DS4 samples [10,26–29]. The quasi-regular eutectics II in DS2 and DS4 samples, with a higher growth rate, shows a higher volume fraction of γ-(Nb, Ti)5Si3 phases compared with the quasi-regular eutectics I in DS1 and DS3 samples. It is abnormal from the point of the degree of segregation. As mentioned above, the growth pattern of (Nb, Ti)5Si3 phases shifts from the faceted growth to the continued growth with the increase in the withdrawal rate in DS2 and DS4 samples. The γ(Nb, Ti)5Si3 phases grow without restrictions. Therefore, larger volume fractions of γ-(Nb, Ti)5Si3 phases exist in the quasi-regular eutectics II. The schematic illustration of eutectic evolution is shown in Fig. 7. The eutectic morphologies are closely related to the withdrawal rate and the contents of Ti and Si. The lower content of Ti and the larger withdrawal rates promote the grow pattern transition of (Nb, Ti)5Si3 phases from faceted growth to continuous growth. The growth rates of (Nb, Ti)5Si3 phases and (Nb, Ti)ss phases are commensurate, contributing to the transformation from quasi-regular eutectics I to quasiregular eutectics II. The higher content of Si and the lower withdrawal rates cause the increase in the degree of segregation, and consequently promote the generation of eutectic dendrites colonies.

Table 2 The volume fractions of different (Nb,Ti)5Si3 phases in DS samples.

DS1 DS2 DS3 DS4

quasi-regular eutectics I quasi-regular eutectics II quasi-regular eutectics I quasi-regular eutectics II

α-Nb5Si3 phases

γ-Nb5Si3 phases

Sum

25.04 ± 1.93%

9.69 ± 1.34%

34.73 ± 0.59%

21.91 ± 1.43%

12.82 ± 0.11%

34.73 ± 1.54%

26.67 ± 0.47%

10.07 ± 0.24%

36.74 ± 0.71%

21.73 ± 1.07%

14.99 ± 1.67%

36.72 ± 0.60%

with higher withdrawal rates (600 μm/s) have been processed and the longitudinal morphologies are shown in Fig. 6. From the longitudinal sections, the microstructures are quasi-regular eutectics II and the dendritic growth of the (Nb, Ti)ss phases is suppressed compared with DS4 and DS6 samples (500 μm/s). It proves that the increase in growth rates promotes the decrease in the degree of segregation and suppresses the dendritic growth of the (Nb, Ti)ss phases.

3.5. The volume fraction of γ-(Nb, Ti)5Si3 phases The γ-(Nb, Ti)5Si3 phases have a great adverse effect on the fracture toughness [46], and therefore the volume fractions of different (Nb,Ti)5Si3 phases in the DS1-DS4 samples are studied and the results are summarized in Table 2. As shown in Table 2, compared with DS3 and DS4 samples, the DS1 and DS2 samples are corresponded to lower contents of Ti, and have less volume fractions sum of γ-(Nb, Ti)5Si3 phases and α-(Nb, Ti)5Si3 phases. It may be due to the change of the

4. Conclusion (1) There were three typical eutectic in Nb–xTi–ySi–4Cr–2Al–2Hf Fig. 7. Schematic illustration of the eutectic evolution. (a) quasiregular eutectics I; (b) quasi-regular eutectics II; (c) eutectic dendrites colonies. (A) the lower content of Ti and the larger withdrawal rates; (B) the higher content of Si; (C) the lower withdrawal rates.

278

International Journal of Refractory Metals & Hard Materials 71 (2018) 273–279

N. Wang et al.

(2)

(3)

(4)

(5)

(22Ti-15Si; 24Ti-15Si; 22Ti-16.5Si, at.%) alloys: quasi-regular eutectics I, quasi-regular eutectics II and eutectic dendrites colonies. The lower content of Ti and larger withdrawal rate contributed to the microstructure transformation from quasi-regular eutectics I to quasi-regular eutectics II. The higher content of Si and lower withdrawal rates led to the increase in the degree of segregation and promoted the generation of eutectic dendrites colonies. The size of the bud-like quasi-regular eutectics I and II increased with the increase of Ti and Si contents. The size of quasi-regular eutectics II are smaller than quasi-regular eutectics I. For the quasi-regular eutectics I and II, the volume fractions of γ(Nb, Ti)5Si3 phases increased with increasing the content of Ti. The growth pattern transition caused more γ-(Nb, Ti)5Si3 phases to exit in the quasi-regular eutectics II compared with the quasi-regular eutectics I.

Alloys Compd. 425 (1) (2006) 223–229. [17] Y. Li, C. Ma, H. Zhang, et al., Mechanical properties of directionally solidified Nb–Mo–Si-based alloys with aligned Nbss/Nb5Si3 lamellar structure, Mater. Sci. Eng. A 528 (18) (2011) 5772–5777. [18] Q. Huang, X. Guo, Y. Kang, et al., Microstructures and mechanical properties of directionally solidified multi-element Nb-Si alloy, Prog. Nat. Sci.: Mater. Int. 21 (2) (2011) 146–152. [19] S. Yuan, L. Jia, L. Ma, et al., Eutectic formation during directional solidification: effect of the withdrawal rate, Mater. Lett. 92 (2013) 317–320. [20] J.F. Weng, L.N. Jia, S.N. Yuan, et al., Microstructure evolution of directionally solidified Nb–Si alloy, Mater. Sci. Technol. 30 (4) (2014) 418–423. [21] Y. Guo, L. Jia, B. Kong, et al., Microstructure transition from lamellar eutectic to anomalous eutectic of Nb-Si based alloy powders by heat treatment and spark plasma sintering, J. Alloys Compd. 696 (2017) 516–521. [22] Z. Xiao, L. Zheng, L. Wang, et al., Microstructure evolution of Ti-47Al-2Cr-2Nb alloy in the liquid-metal-cooling (LMC) directional-solidification process, J. Wuhan Univ. Technol. Mater. Sci. Ed. 26 (2) (2011) 197–201. [23] T.W. Clyne, Heat flow in controlled directional solidification of metals: I. Experimental investigation, J. Cryst. Growth 50 (3) (1980) 684–690. [24] T.W. Clyne, Heat flow in controlled directional solidification of metals: II. Mathematical model, J. Cryst. Growth 50 (3) (1980) 691–700. [25] J.C. Zhao, B.P. Bewlay, M.R. Jackson, et al., Alloying and phase stability in niobium silicide in-situ composites, Struct. Intermetallics (2001) 483–491. [26] S. Mathieu, S. Knittel, M. François, et al., Towards the improvement of the oxidation resistance of Nb-silicides in situ composites: a solid state diffusion approach, Corros. Sci. 79 (2014) 119–127. [27] K. Zelenitsas, P. Tsakiropoulos, Study of the role of Ta and Cr additions in the microstructure of Nb–Ti–Si–Al in situ composites, Intermetallics 14 (6) (2006) 639–659. [28] J. Geng, P.A. Tsakiropoulos, Study of the microstructures and oxidation of Nb–Si–Cr–Al–Mo in situ composites alloyed with Ti, Hf and Sn, Intermetallics 15 (3) (2007) 382–395. [29] I. Grammenos, P. Tsakiropoulos, Study of the role of Al, Cr and Ti additions in the microstructure of Nb–18Si–5Hf base alloys, Intermetallics 18 (2) (2010) 242–253. [30] K. Suganuma, T. Okamoto, M. Koizljmi, et al., Effect of interlayers in ceramic-metal joints with thermal expansion mismatches, J. Am. Ceram. Soc. 67 (12) (1984). [31] L. Zhang, Wu J. Thermal, Expansion and elastic moduli of the silicide based intermetallic alloys Ti5Si3 (X) and Nb5Si3, Scr. Mater. 38 (2) (1997) 307–313. [32] S. Zhang, X. Guo, Effects of B addition on the microstructure and properties of Nb silicide based ultrahigh temperature alloys, Intermetallics 57 (2015) 83–92. [33] I. Barin, Thermochemical Data of Pure Substances, Thermochemical Data of Pure Substances, Wiley-VCH, 1997. [34] Y. Li, S. Miura, K. Ohsasa, et al., Ultrahigh-temperature Nbss/Nb5Si3 fully-lamellar microstructure developed by directional solidification in OFZ furnace, Intermetallics 19 (4) (2011) 460–469. [35] T.G. Chart, Thermochemical data for Transition-Metal-Si systems, High Temp. High Pressures 5 (3) (1973) 241–252. [36] S.D. Peteves, R. Abbaschian, Growth kinetics of solid-liquid Ga interfaces: part I. Experimental, Metall. Trans. A. 22 (6) (1991) 1259–1270. [37] W. Dai, S. Xue, J. Lou, et al., Development of Al–Si–Zn–Sr filler metals for brazing 6061 aluminum alloy, Mater. Des. 42 (2012) 395–402. [38] S. Amini, H. Kalaantari, S. Mojgani, et al., Graphite crystals grown within electromagnetically levitated metallic droplets, Acta Mater. 60 (20) (2012) 7123–7131. [39] G. Liu, L. Liu, C. Ai, et al., Influence of withdrawal rate on the microstructure of Nibase single-crystal superalloys containing Re and Ru, J. Alloys Compd. 509 (19) (2011) 5866–5872. [40] X. Zhao, L. Liu, Z. Yu, et al., Influence of directional solidification variables on the microstructure and crystal orientation of AM3 under high thermal gradient, J. Mater. Sci. 45 (22) (2010) 6101–6107. [41] C. Ai, X. Zhao, L. Liu, et al., Influence of withdrawal rate on last stage solidification path of a Mo-rich Ni3Al based single crystal superalloy, J. Alloys Compd. 623 (2015) 362–366. [42] R. Goetzinger, M. Barth, D.M. Herlach, Growth of lamellar eutectic dendrites in undercooled melts, J. Appl. Phys. 84 (3) (1998) 1643–1649. [43] J.F. Li, Y.H. Zhou, Eutectic growth in bulk undercooled melts, Acta Mater. 53 (8) (2005) 2351–2359. [44] X. Ma, L. Liu, Solidification microstructures of the undercooled Co–24at% Sn eutectic alloy containing 0.5 at% Mn, Mater. Des. 83 (2015) 138–143. [45] X.X. Wei, X. Lin, W. Xu, et al., Remelting-induced anomalous eutectic formation during solidification of deeply undercooled eutectic alloy melts, Acta Mater. 95 (2015) 44–56. [46] K.S. Chan, Alloying effects on the fracture toughness of Nb-based silicides and Laves phases, Mater. Sci. Eng. A 409 (1) (2005) 257–269.

Acknowledgement The authors gratefully acknowledge the financial support by National Natural Science Foundation of China under the contract of 51471013, 51571004 and 51401017. References [1] J.H. Perepezko, The hotter the engine, the better, Science 326 (5956) (2009) 1068–1069. [2] T.M. Pollock, Alloy design for aircraft engines, Nat. Mater. 15 (8) (2016) 809–815. [3] B.P. Bewlay, M.R. Jackson, P.R. Subramanian, et al., A review of very-high-temperature Nb-silicide-based composites, Metall. Mater. Trans. A 34 (10) (2003) 2043–2052. [4] Okamoto H, Phase diagrams for binary alloys, Desk Handbook, 2000 ASM International, Member/Customer Service Center, Materials Park, OH 44073-0002, USA, 2000, p. 828. [5] G.M. Cheng, L.L. He, Microstructure evolution and room temperature deformation of a unidirectionally solidified Nb-22Ti-16Si-3Ta-2Hf-7Cr-3Al-0.2Ho (at.%) alloy, Intermetallics 19 (2) (2011) 196–201. [6] Y. Pan, Y. Lin, H. Wang, et al., Vacancy induced brittle-to-ductile transition of Nb5Si3 alloy from first-principles, Mater. Des. 86 (2015) 259–265. [7] B. Kong, L. Jia, H. Zhang, et al., Microstructure, mechanical properties and fracture behavior of Nb with minor Si addition, Int. J. Refract. Met. Hard Mater. 58 (2016) 84–91. [8] Z. Sun, X. Guo, B. Guo, Effect of B and Ti on the directionally solidified microstructure of the Nb–Si alloys, Int. J. Refract. Met. Hard Mater. 51 (2015) 243–249. [9] Z. Li, L.M. Peng, Microstructural and mechanical characterization of Nb-based in situ composites from Nb–Si–Ti ternary system, Acta Mater. 55 (19) (2007) 6573–6585. [10] Y.X. Tian, J.T. Guo, G.M. Cheng, et al., Effect of growth rate on microstructure and mechanical properties in a directionally solidified Nb-silicide base alloy, Mater. Des. 30 (6) (2009) 2274–2277. [11] L. Su, L. Jia, Y. Feng, et al., Microstructure and room-temperature fracture toughness of directionally solidified Nb–Si–Ti–Cr–Al–Hf alloy, Mater. Sci. Eng. A 560 (2013) 672–677. [12] W. Liu, J.B. Sha, Failure mode transition of Nb phase from cleavage to dimple/tear in Nb-16 Si-based alloys prepared via spark plasma sintering, Mater. Des. 111 (2016) 301–311. [13] W.Y. Kim, H. Tanaka, A. Kasama, et al., Microstructure and room temperature fracture toughness of Nbss/Nb5Si3 in situ composites, Intermetallics 9 (9) (2001) 827–834. [14] W.Y. Kim, H. Tanaka, S. Hanada, Microstructure and high temperature strength at 1773 K of Nbss/Nb5Si3 composites alloyed with molybdenum, Intermetallics 10 (6) (2002) 625–634. [15] N. Sekido, Y. Kimura, S. Miura, et al., Solidification process and mechanical behavior of the Nb/Nb5Si3 two phase alloys in the Nb-Ti-Si system, Mater. Trans. 45 (12) (2004) 3264–3271. [16] N. Sekido, Y. Kimura, S. Miura, et al., Fracture toughness and high temperature strength of unidirectionally solidified Nb–Si binary and Nb–Ti–Si ternary alloys, J.

279