Effect of Sintering Temperature on Properties of SiC Fiber Reinforced Tungsten Matrix Composites

Abstract

Continuous SiC fibers were used to toughen tungsten in this work. The composites were prepared by hot-press using W powders sintered from 1500°C to 1900°C for 1 h with 20 MPa pressure. Phases changes, microstructure, mechanical and thermal properties of composites were examined. Besides, the kinetics between SiC and W was also studied. The stress-strain curve had pseudo-ductility at room temperature, except for 1900°C sintered samples due to complete damage of fiber. Thermal conductivity was calculated from two directions: in-plane and through-plane directions (parallel and vertical to fiber direction), in which through-plane direction display higher thermal conductivity than other direction. Besides, severe reaction was verified between SiC and W, and the reaction rate increased 18 times when temperature increased from 1500°C to 1700°C. Therefore, it is suitable to use SiC fiber to reinforce W to ameliorate the brittleness with limited reaction between fiber and matrix.

Fig. 5 Results of tensile tests of SiC_f/W composites and pure W. Images (a) to (d) show the strain-stress curve of composites sintered from 1500°C to 1900°C respectively; Image (e) shows tensile test result of pure W sintered at 1700°C.

1. Introduction

Tungsten (W) is one of the most potential candidate materials in fusion reactor as a plasma-facing material (PFM) owing to its high melting point (3440°C), high thermal conductivity, and good resistance against plasma sputtering and corrosion, etc. However, W materials, as body-centered cubic (BCC) structure metal, exhibit brittle fracture at room temperature (RT) because of their relatively high ductile-brittle transition temperature (DBTT). Therefore, the application of W materials is greatly restricted. Under this condition, for W with high strength and low toughness, improving its brittle fracture behavior (namely toughening) is the key to enhance its mechanical properties and to satisfy the engineering applications.

Many solutions were investigated to enhance the toughness of W, such as adding alloying elements,^1,2) second phase particles^3–11) and W fibers,^12,13) and machining deformation technology,¹⁴⁾ etc. Re element has been widely added to BCC structural metals such as W, Mo and Cr to improve toughness.²⁾ For W, Re can change the symmetry properties of the 1/2⟨111⟩ screw dislocation core to increase the number of slip surfaces, and on the other hand, it can reduce the Peierls force of plastic deformation, leading to improvement of the mobility of dislocation.¹⁵⁾ Besides, the grain boundary bonding force also can be improved by adding Re.¹⁶⁾ But W–Re alloy exhibit hardening and brittle behavior after neutron irradiation because of irradiation-induced precipitation WRe (σ phase) and WRe₃ (χ phase).^17,18) Various second phases (K,^4,11) TiC,^9,10) ZrC,⁸⁾ SiC³⁾ and La₂O₃,⁵⁾ etc.) were used to reinforce W material by refining the grain size, and display higher strength and better creep property but lower tensile elongation compared with pure W.¹⁹⁾ Thus, this method should be a way to strengthen rather than toughening because the added second phase is dispersed in the grain boundary, which can suppress the movement of dislocation. Processing deformation is an effective and easy way for engineering toughening method, but obtained W can recrystallize after heat treatment (higher than 1300°C),²⁰⁾ thus exhibiting recrystallization brittle behavior.²¹⁾ W fiber toughened tungsten matrix composites are also considered an effective way to improve W toughness, and the bulk W fiber toughened W composites still display pseudo-ductile behavior at RT.²²⁾ However, the degradation of W fiber reinforced composites by neutron irradiation is considered due to the degradation of bulk W by neutron irradiation.

In this work, a new way is provided for designing the structural materials in fusion nuclear systems. The SiC fiber was considered as reinforcement to W because SiC exhibits excellent properties. It has been reported that SiC material with high crystallinity exhibits very excellent properties under neutron irradiation environment and does not cause strength deterioration.²³⁾ Furthermore, the coefficients of thermal expansion of W and ceramic fibers are very close to each other, so the generated stress at the interface by thermal load should be limited. Umer et al.³⁾ synthesized SiC particles reinforced W and found that the flexural strength decreased with the increment of SiC content, while the ablation resistance improved because the layer of SiO₂ appeared on the surface of ablated samples to prevent the formation of WO₃. Shi et al.²⁴⁾ reported 0.8 mass% discontinuous SiC fiber reinforced W–20Cu composites had 1200 MPa by 3-point bending method and thermal conductivity was 235 W/(m·K). Nevertheless, both works did not focus on toughness and did not illustrate the strain-stress curve in the papers, and no works focused on the continuous SiC_f/W system. Thus, discussing the fracture behavior and thermal properties of continuous SiC fiber reinforced W composites is necessary to give more selections for diverter application. In this work, unidirectional SiC fibers as reinforcement were used to fabricate W based composites by hot-press to widen the operating temperature of W limited by high DBTT and recrystallization behavior, and to maintain toughness after irradiation under neutron irradiation or plasma exposure. Then the mechanical property was examined by tensile test at RT to evaluate whether the SiC fiber can enhance the ductility of W. The density, microstructure and reactions in the sintering process were also discussed, as well as the kinetics between SiC and W.

2. Experimental

The schematic diagram about the process of fabricating composites in this work is displayed in Fig. 1. Composites were synthesized by hot-press using continuous unidirectional Hi-Nicalon type S SiC fiber (NGS Advanced Fibers Co., Ltd.), and W powder with 0.6 um grain size (99.9% purity, Kojundo Ltd. Company). To fabricate the composites, W powder and PVB were dispersed in acetone by ball-milling for 24 h using W balls to prepare slurry firstly. Then slurry was poured to SiC fiber bundle, which was desized at 500°C before fiber was used. Subsequently, fibers with slurry were dried at RT for 24 h. The prepreg was cut to the size of 40 × 40 mm², then arranged with the same direction of fiber for sintering at various temperatures from 1500°C to 1900°C for 1 h with 20 MPa pressure in Ar (see Fig. 1). Besides, to understand the kinetics of reactions between SiC and W, the W foil and chemical vapor deposition (CVD) SiC diffusion couples were joined at the same pressure and holding time with the sintering composites, while the temperature was from 1500°C to 1700°C.

Fig. 1

Method of preparation SiC fiber reinforced W composite.

The density ρa and ρs of composite was measured by Shimadzu AccuPyc II 1340 using helium gas and size respectively. Tensile tests were carried out by Instron-5581 at RT to assess the mechanical properties. The size of the testing bar was 40 × 3 × 1.5 mm³. All surfaces were polished, and two pieces of strain gauge were pasted to 3 × 40 mm surface respectively before measurement. The microstructure of composites was observed by scanning electron microscopy (SEM). In this work, electron probe microanalyzer (EPMA) was used for elemental analysis because energy-dispersive X-ray spectroscopy (EDS) cannot distinguish W and Si elements for the reason that Kα emission energy between W and Si is too close. The phase analysis was carried out by X-ray diffraction (XRD) using a Co target. The heat capacity was measured by differential scanning calorimetry (DSC), and thermal diffusivity of composites were tested by light flash apparatus (LFA) using the specimen with a diameter of 6 mm and thickness of 2 mm. Then the thermal conductivity was calculated by the equation of λ = Cp * ρ * a (λ, thermal conductivity; ρ, density; a, thermal diffusivity; Cp, heat capacity).

3. Results

3.1 Density and porosity

The results of density by size and densitometer are given in Table 1. The fiber volume fraction displayed here was calculated by weight and density (3.1 g/cm³). The results show that the densification happened when sintering temperature was increased from 1500°C to 1900°C. In addition, the values of density by these two methods differ greatly, which caused from high open porosity. However, the density by size including open and closed pores is more reasonable for a bulk material.

Table 1 Density of sintered composites.

3.2 Phase changes and microstructure after sintering

Figure 2 reveals the XRD patterns of fabricated composites, implying the effect of sintering temperature on interfacial reaction. Besides, the schematic diagram of the measured surface is also exhibited in the inset of the XRD pattern image, in which the tested surface is parallel to the direction of fiber. Phase compositions of composites changed a lot as a result of reactions between W and SiC. Two different tungsten carbides (W₂C and WC) and two different tungsten silicide phases (W₅Si₃ and WSi₂) were identified after sintering, which is different from the work of SiC powder reinforced W prepared by the SPS method. There is only W₂C and W₅Si₃ found after sintering at 1700°C.³⁾ In this work, the peak of W can only be identified at 1500°C prepared samples because of the small grain size of raw W powder and SiC fiber as well as the relatively large surface of small diameter fiber, bringing about the severe diffusion of Si or C atom, which suggests that the reaction rate enhance a lot by increasing the sintering temperature from 1500°C to 1900°C. Another phenomenon is the peak of WC and W₅Si₃ becoming stronger with the sintering temperature enhancement because W₂C and WSi₂ transferred to WC and W₅Si₃.²⁵⁾ While in previous papers, W₂C always acts as a more stable phase in the W–Si–C ternary system.^26–28)

Fig. 2

XRD pattern of synthesized composites fabricated at different temperature.

Figure 3 exhibits the secondary electron (SE) images and the elements’ distribution measured by EPMA of composites fabricated at 1500°C and 1600°C. Tungsten carbides located near the SiC part, and a gap of Si distribution in matrix can be found in the EPMA results, and a similar phenomenon was observed in the paper,²⁶⁾ indicating that W is easier to react with Si, rather than C. Moreover, in both kinds of composites, the Si and C atoms diffused to the W region because of their smaller atomic radius than the size of W atom, in which, the region of Si atom diffused is larger than the length of C from image (b), suggesting that the reaction rate between Si and W is higher than that of W and C. Moreover, Si atoms have already diffused the whole matrix in composites fabricated above 1500°C, due to the small grain size of W as raw material, and no W remained according to the results of EPMA (see Fig. 3(b)) and XRD. Besides, because the reaction rate increased with temperature according to XRD, therefore the generated phases (tungsten silicides and tungsten carbides) are not appropriate as diffusion barriers in the SiC/W system, which also can be found in the microstructure images of composites.

Fig. 3

SEM images (secondary electron mode) and element analysis by EPMA mapping of composites fabricated at different temperatures. (a) 1500°C; (b) 1600°C.

The morphologies of the cross-section images with different magnifications of SiC_f/W composites sintered from 1500°C to 1900°C after tensile tests examined by SEM in SE mode are shown in Fig. 4. For both composites, tiny pores in matrix near fiber tow region can be observed in W powders synthesized from 1500°C to 1700°C, and pores also can be found in the middle matrix part in 1500°C prepared composite. These are unsintered regions caused by W and carbides with the higher melting point than silicides, combined with the EPMA results. Besides, five layers, namely carbides/silicides/W/silicides/carbides, can be observed at matrix in composites sintered at 1500°C. In comparison, W layer disappeared, and 3 layers (carbides/silicides/carbides) remained from 1600°C. Thus, the pores at the fiber edge region is caused by the existence of carbides with high melting point. Furthermore, in the composite sintered 1900°C, the matrix region is totally dense except for the slurry undispersed region, and fiber damaged completely. Thus, 1900°C is too high to sinter SiC fiber reinforced W composite. Therefore, it is necessary to discover an effective interface as a diffusion barrier to impede the reaction at the high temperature.

Fig. 4

SEM images in secondary electron mode of cross-section of SiC fiber reinforced W composites. (a1), (a2) are the images of composite fabricated at 1500°C with different magnification. (b1), (b2) are the images of composite fabricated at 1600°C with different magnification. (c) is the image of composite sintered at 1700°C. (d) is the image of composite prepared at 1900°C.

3.3 Mechanical property

Figure 5 reveals the variation of tensile strength at RT of SiC_f/W composites fabricated at the different sintering temperatures, in which tensile test data of pure W annealed at 1700°C is also shown for comparison. The load direction was the same as the fiber direction. Furthermore, it can be found that composites sintered at 1500°C to 1700°C exhibited pseudo ductility, compared with pure W annealed at 1700°C (see Fig. 5(e)), although the tensile tests were carried out at RT, which is lower than the DBTT of W.

Fig. 5

Results of tensile tests of SiC_f/W composites and pure W. Images (a) to (d) show the strain-stress curve of composites sintered from 1500°C to 1900°C respectively; Image (e) shows tensile test result of pure W sintered at 1700°C.

Furthermore, the tendency of pseudo ductility reduced with increasing the sintering temperature, and almost no pseudo-ductile behavior can be found in 1900°C sintered samples from the stress-strain curve because of significant reaction between fiber and matrix. Because relatively low sintering temperature can decrease the reaction rate between SiC fiber and W causing more unreacted fibers remained after sintering, which contributed for the higher ductility. In addition, strength increases with the enhancement of temperature. The average ultimate tensile strength (UTS) of composites sintered at 1900°C was about 178 MPa, which is almost three times higher than 56 MPa of 1500°C sintered composites. In addition, the average UTS of composites prepared at 1600°C and 1700°C are 37.8 MPa and 82.9 MPa, respectively. Compared with pure W, the composites showed relatively lower UTS but more apparent pseudo ductility. It is unnecessary to consider the DBTT for fiber reinforced composites because brittleness of W isn’t matter.

3.4 Thermal property

Effect of sintering temperature on thermal conductivity was characterized. Thermal conductivity (λ) of composites measured in two different directions at RT, which is thermal load vertical (in-plane, also named X direction) and parallel (through-plane, also named Y direction) to the fiber direction, respectively, was shown in Fig. 6. Furthermore, the geometric density with open and closed pores was used to calculate λ because of the big difference compared with the value measured by the helium pycnometer (see Table 1). The results show the λ increased with sintering temperature caused by densification. Moreover, it is substantial to be found that the value of λ is higher in the through-plane direction (red points) compared to the in-plane direction (black points). Because thermal conductivity of X direction is determined by mixture of W layer and SiC fiber layer with pores, while it is determined by the highest thermal conductivity layer at y direction. The thermal conductivity of 1700°C fabricated composite in through-plane direction showed the best result of 51.5 W/(m·K) at RT, whereas the value was 32.2 W/(m·K) in the in-plane direction. However, even the highest thermal conductivity of SiC_f/W composites is still lower than the thermal conductivity of pure W of 170 W/(m·K)¹¹⁾ and W-K alloy of 160 W/(m·K)⁴⁾ as well as K doped W–3 mass%Re alloy of 110 W/(m·K),²⁹⁾ while similar with W–10 mass%Re alloy of 60 W/(m·K) and higher than W–25 mass%Re alloy of 25 W/(m·K).³⁰⁾ Because the reaction products display the low conductivity compared to W and SiC, another reason is the low relative density even for 1700°C sintered sample.

Fig. 6

Thermal conductivity of SiC_f/W composites fabricated at 1500°C to 1700°C in different directions (X direction, in-plane direction; Y direction, through-plane direction) measured at 25°C.

3.5 Kinetics based on W/SiC diffusion couples

The W foil and a CVD SiC plate were joined by hot-press to understand the kinetics because it is easy to measure the correct thickness of reaction zone compared with the composites. Figure 7 reveals the cross-sectional SEM images of W/SiC joints. The average thickness of the reaction zone measured from these images from different regions was 7.1 µm for 1500°C, 13.3 µm for 1600°C, and 30.4 µm for 1700°C, respectively. To understand the reaction kinetics better, the following equations were used to calculate the growth rate constant and apparent activation energy.   

\begin{equation} x^{2} = 2k_{p}t \end{equation} (1)

  

\begin{equation} k_{p} = k_{0}\exp (-Q/RT) \end{equation} (2)

in which x is the thickness of reaction zone, k_p is the parabolic growth rate constant, t is time equal to 3600 s for every joint, k₀ is pre-exponential factor, Q is apparent activation energy, R is gas constant 8.314 J/(mol*K), T is absolute temperature with the unit of Kelvin. The parabolic growth rate for W/SiC diffusion couples from 1773 K to 1973 K based by eq. (1) are 7.88 × 10⁻¹⁵ m²/s, 2.76 × 10⁻¹⁴ m²/s, and 1.44 × 10⁻¹³ m²/s, respectively. Therefore, the reaction rate increased approximately 3.5 times when the temperature increased from 1500°C to 1600°C, while it is about 5 times higher at 1700°C than that of 1600°C, matching the XRD results (see Fig. 2), implying that the reaction rate increases with temperature. Thus the generated phases (tungsten silicides and tungsten carbides) are not appropriate as diffusion barriers in the SiC/W system. In addition, the apparent activation energy was found to be equal to 422 kJ/mol, and k₀ is 0.019 m²/s according to Fig. 7(d) from the Arrhenius equation of eq. (2).

Fig. 7

The cross-sectional SEM images of the W/SiC diffusion couple joined at different temperatures for 1 h. (a) 1500°C; (b) 1600°C; (c) 1700°C; and (d) is for Arrhenius plots of the parabolic rate constants of the reaction zones by the W/SiC joints as a function of the reciprocal of temperature (K).

4. Discussion

Table 2 summarizes the mechanical and thermal properties of synthesized composites. The reaction between SiC fiber and W matrix depended on sintering temperature. Only limited reactions happened in relatively lower sintering temperature (1500°C), although more than half of W was reacted. For mechanical property, the decreasement of UTS from 1500°C to 1600°C is from the reduction of high strength W ratio even with higher densification. Therefore, reactivity has more contribution for UTS. Sufficient reactions between SiC and W were observed above 1600°C. The strength increased with sinterability above 1600°C. In addition, for composite sintered at lower temperature, short pull-out fiber can be observed. Moreover, pores at the fiber bundle and the fiber edge caused the weak interface between fibers and matrix, so the laminate effect is also responsible for the pseudo-ductility. However, densification lead to the strong interface between fiber and matrix, causing that the laminate effect reduced with increasing the sintering temperature. Therefore, pseudo ductile behavior was limited with the sintering temperature so that no pseudo-ductility exists in composite fabricated at 1900°C with the highest density and the most severe reaction. So, densification has a greater impact on the pseudo-ductility. While for the thermal conductivity, 1500°C sintered composite shows the lowest thermal conductivity even if W with high thermal conductivity exists in the matrix due to low sinterability, and it increases with the sintering temperature. The ρa in the Table 1 was calculated without open pores. A lot of open pores existed in the material sintered at 1500°C. Therefore, sinterability dominates the thermal conductivity compared with reactivity. Thus, there is a tradeoff. Relatively high temperature sintering improves densification, strength and thermal conductivity, however high temperature caused severer damage in fiber and matrix. So the densification of composites and reaction between fiber and matrix need to be balanced. 1700°C should be suitable temperature to fabricate composites, which displayed pseudo ductile behavior with relative high UTS, as well as high thermal conductivity, however, the composition of matrix has been reacted totally, and no W remained. Based on this, to raise the pseudo-ductility and thermal conductivity of the composites, it is essential to prepare diffusion barrier to avert the reaction to remain higher content of W and keep the shape of SiC in composites after sintering.

Table 2 Summary of the fabricated composites’ properties.

5. Conclusion

SiC fiber reinforced W composites were fabricated at different sintering temperature by hot-pressing successfully. The strength and thermal conductivity of composites increased with increment of sintering temperature due to densification, however pseudo ductility was reduced due to enhanced reaction between SiC fiber and W matrix. In addition, the stress-strain curves displayed obvious pseudo-ductile behavior for specimen sintered at 1500°C, 1600°C and 1700°C compared with pure W annealed at 1700°C even if the temperature of tensile test is lower than the DBTT of W. So SiC fiber as reinforcement to strengthen W is feasible. Concerning thermal conductivity, higher sintering temperature leaded to higher thermal conductivity. In addition, it is much higher in through-plane direction than in-plane direction. The composites fabricated at 1700°C showed the largest strength with pseudo ductility. Besides, the reaction rate increased approximately 18 times when temperature increased from 1500°C to 1700°C. Thus, to acquire better properties, it is necessary to recognize an effective diffusion barrier for the W and SiC system without reactions to get higher W content and protect SiC fiber.

Acknowledgement

This work was supported by JSPS KAKENHI Grant Number 21H01063.

REFERENCES

• 1)   Y.  Mutoh,  K.  Ichikawa,  K.  Nagata and  M.  Takeuchi: J. Mater. Sci. 30 (1995) 770–775. doi:10.1007/BF00356341
• 2)  W.D. Klopp, United States: N. p., 1968. osti.gov/biblio/5785668.
• 3)   M.A.  Umer,  D.  Lee,  O.A.  Waseem,  H.J.  Ryu and  S.H.  Hong: Met. Mater. Int. 22 (2016) 493–500. doi:10.1007/s12540-016-5700-y
• 4)   B.  Huang,  L.  Chen,  W.  Qiu,  X.  Yang,  K.  Shi,  Y.  Lian,  X.  Liu and  J.  Tang: J. Nucl. Mater. 520 (2019) 6–18. doi:10.1016/j.jnucmat.2019.03.056
• 5)   M.  Mabuchi,  K.  Okamoto,  N.  Saito,  M.  Nakanishi,  Y.  Yamada,  T.  Asahina and  T.  Igarashi: Mater. Sci. Eng. A 214 (1996) 174–176. doi:10.1016/0921-5093(96)10377-4
• 6)   X.Y.  Zhu,  J.  Zhang,  L.M.  Luo,  J.  Chen,  J.G.  Cheng and  Y.C.  Wu: Adv. Powder Technol. 26 (2015) 640–643. doi:10.1016/j.apt.2015.01.015
• 7)   H.  Kurishita,  S.  Matsuo,  H.  Arakawa,  T.  Sakamoto,  S.  Kobayashi,  K.  Nakai,  T.  Takida,  M.  Kato,  M.  Kawai and  N.  Yoshida: J. Nucl. Mater. 398 (2010) 87–92. doi:10.1016/j.jnucmat.2009.10.015
• 8)   R.  Liu,  Z.M.  Xie,  J.F.  Yang,  T.  Zhang,  T.  Hao,  X.P.  Wang,  Q.F.  Fang and  C.S.  Liu: Nucl. Mater. Energy 16 (2018) 191–206. doi:10.1016/j.nme.2018.07.002
• 9)   Y.  Ishijima,  H.  Kurishita,  H.  Arakawa,  M.  Hasegawa,  Y.  Hiraoka,  T.  Takida and  K.  Takebe: Mater. Trans. 46 (2005) 568–574. doi:10.2320/matertrans.46.568
• 10)   Y.  Kitsunai,  H.  Kurishita,  H.  Kayano,  Y.  Hiraoka,  T.  Igarashi and  T.  Takida: J. Nucl. Mater. 271–272 (1999) 423–428. doi:10.1016/S0022-3115(98)00753-3
• 11)   S.  Nogami,  W.  Guan,  A.  Hasegawa and  M.  Fukuda: Fus. Sci. Technol. 72 (2017) 673–679. doi:10.1080/15361055.2017.1347463
• 12)   Y.  Mao,  J.W.  Coenen,  J.  Riesch,  S.  Sistla,  J.  Almanstötter,  B.  Jasper,  A.  Terra,  T.  Höschen,  H.  Gietl,  C.  Linsmeier and  C.  Broeckmann: Compos. Part A 107 (2018) 342–353. doi:10.1016/j.compositesa.2018.01.022
• 13)   J.W.  Coenen,  Y.  Mao,  S.  Sistla,  J.  Riesch,  T.  Hoeschen,  C.  Broeckmann,  R.  Neu and  C.  Linsmeier: Nucl. Mater. Energy 15 (2018) 214–219. doi:10.1016/j.nme.2018.05.001
• 14)   S.  Miao,  Z.M.  Xie,  X.D.  Yang,  R.  Liu,  R.  Gao,  T.  Zhang,  X.P.  Wang,  Q.F.  Fang,  C.S.  Liu,  G.N.  Luo,  X.  Liu and  Y.Y.  Lian: Int. J. Refract. Met. Hard Mater. 56 (2016) 8–17. doi:10.1016/j.ijrmhm.2015.12.004
• 15)   L.  Romaner,  C.  Ambrosch-Draxl and  R.  Pippan: Phys. Rev. Lett. 104 (2010) 195503. doi:10.1103/PhysRevLett.104.195503
• 16)   W.  Setyawan and  R.J.  Kurtz: Scr. Mater. 66 (2012) 558–561. doi:10.1016/j.scriptamat.2012.01.002
• 17)   Y.  Nemoto,  A.  Hasegawa,  M.  Satou and  K.  Abe: J. Nucl. Mater. 283–287 (2000) 1144–1147. doi:10.1016/S0022-3115(00)00290-7
• 18)   A.  Hasegawa,  M.  Fukuda,  S.  Nogami and  K.  Yabuuchi: Fusion Eng. Des. 89 (2014) 1568–1572. doi:10.1016/j.fusengdes.2014.04.035
• 19)   M.  Rieth and  B.  Dafferner: J. Nucl. Mater. 342 (2005) 20–25. doi:10.1016/j.jnucmat.2005.03.013
• 20)   K.  Tsuchida,  T.  Miyazawa,  A.  Hasegawa,  S.  Nogami and  M.  Fukuda: Nucl. Mater. Energy 15 (2018) 158–163. doi:10.1016/j.nme.2018.04.004
• 21)   C.  Ren,  Z.Z.  Fang,  L.  Xu,  J.P.  Ligda,  J.D.  Paramore and  B.G.  Butler: Acta Mater. 162 (2019) 202–213. doi:10.1016/j.actamat.2018.10.002
• 22)   H.  Gietl,  S.  Olbrich,  J.  Riesch,  G.  Holzner,  T.  Höschen,  J.W.  Coenen and  R.  Neu: Eng. Fract. Mech. 232 (2020) 107011. doi:10.1016/j.engfracmech.2020.107011
• 23)   T.  Koyanagi,  K.  Ozawa,  T.  Hinoki,  K.  Shimoda and  Y.  Katoh: J. Nucl. Mater. 448 (2014) 478–486. doi:10.1016/j.jnucmat.2013.10.005
• 24)   X.  Shi,  M.  Wang,  S.  Zhang and  Q.  Zhang: Int. J. Refract. Met. Hard Mater. 41 (2013) 60–65. doi:10.1016/j.ijrmhm.2013.02.002
• 25)   S.  Coşkun,  M.L.  Öveçoğlu,  B.  Özkal and  M.  Tanoğlu: J. Alloy. Compd. 492 (2010) 576–584. doi:10.1016/j.jallcom.2009.11.185
• 26)   S.J.  Son,  K.H.  Park,  Y.  Katoh and  A.  Kohyama: J. Nucl. Mater. 329–333 (2004) 1549–1552. doi:10.1016/j.jnucmat.2004.04.285
• 27)   F.  Goesmann and  R.  Schmid-Fetzer: Mater. Sci. Eng. B 34 (1995) 224–231. doi:10.1016/0921-5107(95)01311-3
• 28)   W.F.  Seng and  P.A.  Barnes: Mater. Sci. Eng. B 76 (2000) 225–231. doi:10.1016/S0921-5107(00)00457-8
• 29)   M.  Fukuda,  A.  Hasegawa and  S.  Nogami: Fusion Eng. Des. 132 (2018) 1–6. doi:10.1016/j.fusengdes.2018.04.117
• 30)   T.  Tanabe,  C.  Eamchotchawalit,  C.  Busabok,  S.  Taweethavorn,  M.  Fujitsuka and  T.  Shikama: Mater. Lett. 57 (2003) 2950–2953. doi:10.1016/S0167-577X(02)01403-9
• 31)   J.  Roger,  F.  Audubert and  Y.  Le Petitcorps: Adv. Eng. Mater. 11 (2009) 399–407. doi:10.1002/adem.200800354