Microstructural Evolution and Mechanical Properties of a Three-Phase Alloy in the Cr–Mo–Nb System

Abstract

A new three-phase alloy of 50Cr–30Mo–20Nb (at%) was studied based on the Cr–Mo–Nb ternary phase diagram, and was composed of a Cr-rich BCC₁ phase, a Mo-rich BCC₂ phase and a NbCr₂ Laves phase after heat treatment at around 1523 K or lower. A supersaturated BCC single-phase solid solution alloy obtained by homogenization at 1973 K for 1 h underwent microstructural evolution during heat treatment at 1473 K. Intragranular precipitation of the Cr-rich BCC₁ phase occurred, which led to the formation of an alternating BCC₁/BCC₂ two-phase microstructure through a discontinuous precipitation process, followed by precipitation of the Laves phase at the BCC₁/BCC₂ interphase boundaries. At a higher temperature of 1523 K, a similar microstructure was observed, with increased BCC decomposition and Laves precipitation rates, while the alloy consisted of BCC and Laves phases at 1773 K. The mechanical properties of alloys heat treated at 1473 K for various periods after the solid-solution treatment were also investigated. A maximum fracture strength of 1493 MPa and a minimum hardness of 773 ± 7 HV were obtained for an alloy aged for 24 h, where the BCC₁/BCC₂ two-phase microstructure dominated. The Vickers hardness of an alloy aged for 72 h, which had a fine-grained microstructure that included the Laves phase, was 839 ± 8 HV under a load of 0.5 kgf, and no obvious microcracks were observed.

Fig. 8 SEM images of 50Cr–30Mo–20Nb alloy after Vickers hardness test: (a) aged at 1473 K for 168 h without solution treatment, (b) aged at 1473 K for 72 h following solution treatment at 1973 K for 1 h. (c), (d) area indicated in (a), and (e), (f) area indicated in (b).

1. Introduction

Refractory metals and their intermetallic compounds have high melting points and excellent high-temperature strength,^1,2) and are considered to be good candidates as structural materials for turbine blades to increase the operating temperature of gas turbines. Some AB₂-type intermetallic Laves phases with topologically close-packed (TCP) structures are classified into such an intermetallic compound group.^3,4) There are three major types of Laves phases: the cubic C15 phase, the hexagonal C14 phase and the dihexagonal C36 phase. However, the limitation of these intermetallics is their poor room-temperature toughness. Ductile phase toughening has been shown to be effective for alleviating the brittleness of Laves phases.^5–8) A larger volume fraction of ductile phases results in tougher materials at room temperature; however, the strength of the ductile phase, mostly the BCC structure for refractory-metal based materials, is not adequate at high temperatures.

To overcome this trade-off, a new BCC₁/BCC₂/Laves three-phase alloy is proposed. The intent is to introduce a second BCC phase with a fine-grained structure to introduce BCC₁/BCC₂ interfaces into the conventional ductile phase. The strength of the combined BCC₁ and BCC₂ phases is expected to be higher than that of a single ductile BCC phase, and the high-temperature strength and creep resistance may be improved by such a microstructural modification, similar to the case for Ni-based γ/γ′ alloys.^9,10) Here, BCC₁ and BCC₂ are Cr-rich and Cr-poor BCC phases, respectively. The authors have previously investigated the crystallographic orientation relationship (OR) between the C15 Laves phase NbCr₂ and the BCC matrix phase in Cr–Mo–Nb alloys with various compositions by considering the lattice mismatch between the C15 Laves phase and the BCC matrix phase.¹¹⁾ NbCr₂ tends to have better lattice compatibility with the Cr-rich BCC₁ phase than with the (Cr-poor) BCC₂ phase. Precipitation of the Laves phase in the BCC₁/BCC₂/Laves three-phase alloy is therefore considered to be governed by the lattice compatibility between the Laves phase and the Cr-rich BCC₁ phase, which results in the formation of fine precipitates. However, the precipitation mechanism is not yet clearly understood.

The purpose of this work was to determine the details of the microstructural evolution mechanism for a BCC₁/BCC₂/Laves three-phase alloy with a composition of 50Cr–30Mo–20Nb. An isothermal section of the Cr–Mo–Nb ternary phase diagram at 1473 K is shown in Fig. 1.¹¹⁾ Microstructural evolution of the 50Cr–30Mo–20Nb alloy at a temperature of 1473 K was mainly considered here, although isothermal treatments at other temperatures were also conducted to study the precipitation of the Laves phase. The strength of alloys heat treated at 1473 K for various periods after solid solution treatment was also evaluated using room-temperature Vickers hardness and compression tests, in order to determine the relationship between the microstructure and crack formation.

Fig. 1

Isothermal section of Cr–Mo–Nb phase diagram at 1473 K.¹¹⁾

2. Experimental Procedure

The purities of the raw materials were 99.9 mass% Nb, 99.99 mass% Cr, and 99.95 mass% Mo. Alloy ingots (ca. 7 g) with a nominal composition of 50Cr–30Mo–20Nb (at%) were prepared by non-consumable arc melting in a high-purity Ar atmosphere on a water-cooled copper hearth. The ingots were turned over and remelted more than ten times to ensure chemical homogeneity. The ingots were solution treated at 1973 K for 1 h in an evacuated high-frequency induction furnace, and then aged at various temperatures (1423, 1473 and 1523 K) for periods up to 100 h. Some of the ingots were only heat treated at 1773 K without solution treatment.

The microstructure was observed using field-emission scanning electron microscopy (FE-SEM; JEOL JXA-8530F) and transmission electron microscopy (TEM; JEOL JEM-2010). The foil specimens for TEM observations were prepared using focused ion beam scanning electron microscopy (FIB-SEM; JEOL JIB-4601F). Selected area electron diffraction (SAED) was employed to determine the crystal structure of the constituent phases. An energy-dispersive X-ray spectroscopy (EDS) system installed in the TEM apparatus was employed to examine the composition of each phase. Image analysis software (Azo-kun, Asahi Kasei Engineering Corporation, Tokyo, Japan) was used to evaluate the areal fractions of the BCC₁, BCC₂ and Laves phases.

The strength of the alloys heat treated at 1473 K for various periods after solid solution treatment was evaluated at room temperature by Vickers hardness testing (ten indents) under a load of 0.5 kgf with a holding time of 30 s, and compression tests were performed with an initial strain rate of 1 × 10⁻⁴ s⁻¹. The Vickers hardness (HV) was determined using:   

\begin{equation} \mathrm{HV} = \frac{F}{A}\approx 1.854\frac{F}{d^{2}}, \end{equation} (1)

where F [kgf] is the applied force, A [mm²] is the surface area of the resulting indentation, and d [mm] is the average length of the diagonal produced by the indenter. The trimmed mean was adopted to evaluate the hardness. The indents were examined using FE-SEM to identify the microcrack formation mechanism. Parallelepiped samples for compression testing with dimensions of 3 × 3 × 6 mm were cut from the aged ingots with a wheel cutter and were mirror polished. Nanoindentation tests (Hysitron TI950, TriboIndenter) were also conducted (fifty indents in total) for samples polished with a cross-section polisher (JEOL SM-09010) to assess the hardness of the constituent phases under a load of 4000 µN.

3. Results and Discussion

3.1 Microstructural evolution following aging at 1473 K

Figures 2 and 3 show the microstructure of the 50Cr–30Mo–20Nb alloy before and after aging at 1473 K for various time periods following solution treatment at 1973 K for 1 h. The phases in the regions indicated by ①–⑤ in Fig. 3 were identified by TEM and EDS, and the results are listed in Table 1. As shown in Fig. 2(a), before aging, the alloy has a single-phase microstructure, with no evidence of NbCr₂ Laves phase precipitates. After aging at 1473 K for 3 h (Fig. 2(b)), dark Cr-rich BCC₁ particles appear in the matrix. At grain boundaries (GBs), the Cr-rich particles are larger and the Mo-rich BCC₂ phases (bright regions) appear. Figure 2(c) shows that the original BCC phase decomposes into dark Cr-rich BCC₁ phases and bright Mo-rich BCC₂ phases in the grain interiors (indicated by white arrows) after aging for 12 h. These phases form a broad banded structure (about 2 µm in width) with circular sub-structures that spread in the grain interiors and terminate at the original BCC phase GBs. The original matrix BCC phase gradually disappears with increasing aging time and is replaced by a BCC₁/BCC₂ two-phase alternating microstructure (Fig. 2(d)). Precipitation and growth of the blocky gray Laves phase occurs following longer aging times, as shown in Figs. 2(e) and 2(f); this phase also appears at GBs in the original BCC phase (area outlined by square in Fig. 2(c)) during the early aging stage. This sequence of microstructural evolution is consistent with previously reported results.¹¹⁾

Fig. 2

Microstructural evolution of 50Cr–30Mo–20Nb alloy before and after aging at 1473 K following solution treatment at 1973 K for 1 h: (a) before aging, and aged for (b) 3 h, (c) 12 h, (d) 24 h, (e) 72 h, and (f) 100 h.

Fig. 3

TEM analyses of alloys aged at 1473 K for (a) 3 h, (b) 6 h, (c, d) 12 h, and (g, h) 100 h. SAED patterns for (e) region A in panel (c), (f) region B in panel (c), (i) region SA in panel (g), and (j) region SA in panel (h). (Numbers refer to regions analyzed using EDS, the results of which are shown in Table 1).

Table 1 EDS results for regions indicated in Fig. 3(a, g, h) (at%).

The TEM images and SAED patterns in Fig. 3 show the details of the microstructure and phases present under different aging conditions. A comparison of Fig. 3(a) and 3(b) shows that the original BCC phase decomposes into BCC₁ and BCC₂ phases following 6 h of aging. From the EDS results shown in Table 1, the black intragranular precipitates in Fig. 2(b) correspond to the Cr-rich BCC₁ phase. This stimulates the subsequent precipitation of a Mo-rich BCC₂ phase in adjacent regions (bright regions), which in turn induces further precipitation of a Cr-rich BCC₁ phase (dark regions), as shown in Fig. 3(b).

Figures 3(c)–(f) show TEM images of the alloy aged for 12 h, and SAED patterns for the original supersaturated BCC structure and the BCC₁/BCC₂ structure, which correspond to the original BCC area (A in Fig. 3(c)) and the BCC₁/BCC₂ two-phase area (B in Fig. 3(c), and the magnified view in Fig. 3(d)), respectively. Diffraction spot doubling is seen in the SAED pattern in Fig. 3(f), which indicates BCC regions with the same crystallographic orientation as each other (cube-on-cube), as reported in a previous paper.¹¹⁾ This verifies the decomposition of the original BCC phase into BCC₁ and BCC₂ phases in the grain interiors, and that the latter phases dominate the microstructure in the early stages of evolution.

TEM images of the alloy aged at 1473 K for 100 h and the corresponding SAED patterns are shown in Fig. 3(g)–(j). The BCC₁ and BCC₂ phases consistently maintain a cube-on-cube relationship, even after long-term aging, as shown in Fig. 3(i). Table 1 also shows that the dark phase is the Mo-rich BCC₂ phase, and the relatively large gray precipitate (denoted as ⑤) in Fig. 3(h) is the NbCr₂ Laves phase. The phase surrounding the Laves phase is mostly the gray Mo-rich BCC₂ phase, which implies that the Laves phase forms at the boundary of the Cr-rich BCC₁ phase and grows by consuming it.

It should be noted that an XRD analysis¹¹⁾ of an alloy aged for 12 h revealed distinct peaks that corresponded to three BCC phases. The lattice constants were different from each other, which strongly suggests that microstructural evolution occurred through discontinuous precipitation (DP), with no evidence for a spinodal-like process. The TEM observations clearly indicate that the BCC₁ and BCC₂ phases are formed by a DP process, and that the precipitation products gradually replace the parent phase (original BCC) by migration of the reaction front. However, this is quite different from typical DP,^12,13) where precipitation typically occurs at GBs, and growth is accompanied by GB migration.¹⁴⁾ Nevertheless, in the present study, intragranular nucleation of the Cr-rich BCC₁ phase in this alloy dominates the early stage of the precipitation process, and becomes a trigger for precipitation of the neighboring Mo-rich BCC₂ phase, which in turn induces further precipitation of the Cr-rich BCC₁ phase, until an alternating BCC₁/BCC₂ two-phase microstructure is finally formed. The mechanism for this type of DP is as yet unclear. This morphology shows a resemblance to spinodal decomposition^15–17) and surface-directed spinodal decomposition.¹⁸⁾ Further investigation of the reasons for the formation of such a unique microstructure is still required.

We previously reported¹¹⁾ three different crystallographic ORs between the BCC matrix phase and the Laves phase precipitates, and it was shown that the lattice mismatch (δ) between the matrix BCC phase and the C15 Laves phase plays an important role in determining the OR. It was concluded that no Laves phase precipitates are formed in the grain interiors during the early stages, due to the relatively large lattice mismatch with the original BCC phase for any OR.

Figures 3(i) and 3(j) show SAED patterns for the BCC and C15 Laves phases, respectively, obtained at the same sample tilt angle. The zone axes for the BCC and C15 phases are ⟨111⟩_BCC and ⟨112⟩_C15, respectively. The {110} reflections for the BCC phase coincide with the {111} reflections for the C15 phase. Thus, the OR between the BCC matrix and the C15 phase is:   

\begin{align*} &(01\bar{1})_{\text{BCC}}\parallel (1\bar{1}1)_{\text{C15}} \\ &[\bar{1}11]_{\text{BCC}}\parallel [\bar{1}12]_{\text{C15}}. \end{align*}

Previous studies^19,20) showed that the OR between the BCC and C15 Laves phases is:   

\begin{align*} &(01\bar{1})_{\text{BCC}}\parallel (1\bar{1}1)_{\text{C15}} \\ &[011]_{\text{BCC}}\parallel [\bar{1}01]_{\text{C15}}, \end{align*}

which was defined as OR 1¹¹⁾ in alloy #2, which corresponds to the present alloy.

The two ORs share common parallel planes, although the directions are different. The angle between the [$\bar{1}11$]_BCC and [011]_BCC zones is 35.26°, and the angle between the [$\bar{1}12$]_C15 and [$\bar{1}01$]_C15 zones is 30°. When the [$\bar{1}11$]_BCC zone is parallel to the [$\bar{1}12$]_C15 zone, the angle between the [011]_BCC and [$\bar{1}01$]_C15 zones is 5.26°. Therefore, the OR observed here can be considered to have a small deviation from OR 1 reported by other researchers and in our previous paper.¹¹⁾ The low δ value between the Cr-rich BCC₁ and Laves phase facilitates the precipitation of the Laves phase by consumption of the Cr-rich BCC₁ phase. As reported by Cui et al.²¹⁾ and Xu et al.,²²⁾ the growth of a (Fe,Cr)₂(Mo,W) type Laves phase close to a (Fe,Cr)₂₃C₆ type M₂₃C₆ carbide phase proceeds by consumption of the Cr-rich carbide in the vicinity. Similarly, in the present study, the NbCr₂ Laves phase surrounded by the Mo-rich BCC₂ phase may grow at the expense of the Cr-rich BCC₁ phase in the vicinity. This is also consistent with previous SEM results.¹¹⁾ A schematic diagram of the proposed microstructural evolution mechanism is shown in Fig. 4.

Fig. 4

Schematic diagram of microstructural evolution of 50Cr–30Mo–20Nb alloy.

3.2 Temperature dependence of microstructural evolution

Figure 5 shows SEM micrographs of specimens aged at three different temperatures. Similar morphologies are obtained for alloys heat treated at 1423, 1473 and 1523 K after solution treatment at 1973 K for 1 h. At relatively high temperature, precipitation of the Laves phase (gray phase) occurs in BCC₁/BCC₂ regions before full BCC decomposition takes place, due to the higher diffusion rate at the BCC₁/BCC₂ interphase boundaries. On the other hand, for heat treatment at 1773 K without a solution treatment, no BCC decomposition is observed and the microstructure indicates a BCC/Laves two-phase alloy.

Fig. 5

SEM images of 50Cr–30Mo–20Nb alloy aged (a) at 1423 K for 24 h after solution treatment (1973 K/1 h), (b) at 1473 K for 24 h after solution treatment, (c) at 1523 K for 12 h after solution treatment, and (d) at 1773 K for 12 h without solution treatment.

The decomposed fraction of the original BCC phase and the fraction of the precipitated Laves phase as a function of the isothermal aging time are shown in Fig. 6. Although the BCC decomposition curves include the Laves phase precipitated in the BCC₁/BCC₂ two-phase structure, they also present the precipitation rates for alternating BCC₁ and BCC₂ phases from the original BCC phase, because the Laves phase always precipitates at BCC₁/BCC₂ interfaces and not in BCC grain interiors. The three BCC decomposition curves (precipitation of BCC₁ and BCC₂ phases) appear to be sigmoidal. The interface area (reaction front) between the decomposition products and the original BCC phase first increases, and then decreases due to the reduction in the amount of the original BCC phase as the aging time is increased.

Fig. 6

Transformed volume fraction for 50Cr–30Mo–20Nb alloy as function of aging time at various temperatures.

The fraction of the Laves phase also increases at a faster rate in the alloy heat treated at higher temperature. Although the fraction of the Laves phase and the aging time appear to have a linear relationship, a similar growth rate is evident at all three temperatures once nucleation has begun. Precipitation of the Laves phase always occurs after the decomposition of the original BCC phase. It starts as soon as individual BCC₁/BCC₂ two-phase areas are formed, so that each area undergoes precipitation for different lengths of time. The Mo-rich BCC₂ phase surrounding the Laves phase may also govern the size of the Laves phase.

Figure 7 shows a time-temperature-transformation (TTT) diagram obtained from the transformation curves. The time taken to achieve 1% transformation to the BCC₁/BCC₂ mixed phase and precipitation of the Laves phase is taken to be the starting time. The transformation curve for Laves phase precipitation is on the right side of the BCC₁/BCC₂ transformation curve because precipitation of the Laves phase always occurs after BCC decomposition. The incubation times for BCC decomposition at 1473 and 1523 K are so close that the ‘nose’ of the curve is likely to be around 1523 K, where the maximum transformation rate is obtained. At higher temperatures, the driving force for the transformation gradually deceases, which results in a lower transformation rate. Although the driving force for precipitation of the BCC₁ and BCC₂ phases continues to increase below the ‘nose’ temperature, the transformation is impeded by slow diffusion. Therefore, two ‘C’ shaped curves can be expected for BCC₁/BCC₂ and Laves phase transformations.

Fig. 7

TTT diagrams for BCC decomposition and Laves phase precipitation in 50Cr–30Mo–20Nb alloy.

3.3 Mechanical properties

Figure 8 shows SEM images of the microstructure of alloys produced under different heat treatment conditions after Vickers indentation tests. In both cases, the area of the indentation is almost the same. For the alloy directly annealed at 1473 K without solution treatment at 1973 K, cracks are present in the extensive Laves phase at the corners of the Vickers indentations because of stress concentration, as shown in Fig. 8(c) and 8(d). The cracks are impeded at the interface between the Laves phase and the Mo-rich BCC₂ phase, which indicates that ductile phase toughening is effective for suppressing crack propagation.^23,24) Zhu et al.²⁵⁾ evaluated the fracture toughness of the NbCr₂ Laves phase, which is around 0.9 MPa$\sqrt{m} $, by Vickers indentation testing using:   

\begin{equation} \mathrm{K} = 0.016\left(\frac{E}{\mathit{HV}}\right)^{1/2}\frac{P}{c^{3/2}}, \end{equation} (2)

where E is Young’s modulus, HV is the Vickers hardness, P is the indentation load and c is the average length of the four surface radial cracks from the indent center to the crack tip. Although it is difficult to estimate the fracture toughness of the present multiphase alloys by Vickers indentation tests because no halfpenny cracks can be produced, even at higher loads,²⁶⁾ the average length of the four radial cracks evaluated according to the previously reported toughness is about 60 µm,^25,27) which is much longer than even the longest crack in the present study (about 20 µm), as shown in Fig. 8. No obvious cracks are apparent at the corners of the indentation in the alloy with a fine-grained microstructure (Fig. 8(e) and 8(f)). The fine-grained microstructure, which is composed of a brittle Laves phase and a BCC₁/BCC₂ two-phase matrix undoubtedly plays an important role in preventing the formation of cracks, which also illustrates that improved fracture toughness could be expected from such alloys with fine-grained structures.

Fig. 8

SEM images of 50Cr–30Mo–20Nb alloy after Vickers hardness test: (a) aged at 1473 K for 168 h without solution treatment, (b) aged at 1473 K for 72 h following solution treatment at 1973 K for 1 h. (c), (d) area indicated in (a), and (e), (f) area indicated in (b).

Figure 9(a) shows the dependence of the hardness on the aging time at 1473 K. A maximum value of 839 ± 8 HV is obtained after 72 h, and this is very similar to the hardness values reported for the NbCr₂ Laves phase.^25,28) Even when it contains the two BCC phases, the alloy still maintains a very high hardness and an improved toughness. Unlike the typical aging process,^29–31) the hardness changes only slightly at the early stage and then reaches a minimum of 773 ± 7 HV after 24 h, when the microstructure is composed of the two BCC phases. Figure 9(b) shows that the hardness of the Cr-rich BCC₁ phase (7.5 GPa) obtained by nanoindentation testing is lower than that for the original supersaturated BCC phase (8.8 GPa) and that for the BCC₂ phase (9.6 GPa). This could therefore be the reason for the decrease in the Vickers hardness during BCC decomposition. The hardness of the alloy is then increased by subsequent aging at 1473 K due to precipitation and growth of hard NbCr₂. The hardness of 14.5 GPa obtained in the present study is much larger than that for a monolithic NbCr₂ Laves phase evaluated by Vickers hardness tests.^25,28) During Vickers hardness measurements of brittle materials, cracks may be formed, and this results in lower hardness values. The hardness of the previously evaluated Laves phase could be lower because of the difference in measurement methods.

Fig. 9

(a) Vickers hardness for 50Cr–30Mo–20Nb alloy as function of aging time at 1473 K, and (b) hardness of each phase measured by nanoindentation.

Figure 10(a) shows compressive stress-strain curves for the 50Cr–30Mo–20Nb alloy aged at 1473 K for various times and tested at room temperature. Even though the alloy with NbCr₂ is toughened by the BCC₁/BCC₂ two-phase matrix, no obvious plastic deformation is observed for any of the alloys at room temperature. A sharp drop in stress (indicated by broken circles) appears in all stress-strain curves during compression, and the compression test for one of the samples was terminated when microcracks were formed (Fig. 10(c)); these were observed using optical microscopy of the surface of the alloy aged for 72 h after unloading. The unalloyed NbCr₂ has been reported to be deformable at temperatures above 1473 K,³²⁾ and the deformation temperature was reduced by the addition of Mo to NbCr₂.³³⁾ In addition, duplex alloys composed of the NbCr₂ Laves phase and the Cr-rich BCC phase showed a significantly reduced brittle-to-ductile transition temperature (BDTT) in the Cr–Nb–Mo system; however, plastic deformation was only observed at temperatures above 1223 K with a large volume fraction of NbCr₂.⁸⁾ Li et al.³⁴⁾ observed a yield stage in arc-melted Cr–20Nb BCC/Laves two-phase alloys during compression testing at room temperature. A subsequent high-temperature observation is planned in order to determine how the BCC phases affect the mechanical properties in terms of the BDTT behavior.

Fig. 10

(a) Compressive stress-strain curves for 50Cr–30Mo–20Nb alloy aged at 1473 K, (b) enlarged view of region where sharp stress drop occurred, and (c) surface microstructure of 1473 K/72 h sample after unloading when sharp stress drop occurred.

The crack initiation strength and fracture strength of each alloy evaluated from compression tests are plotted in Fig. 11. The fracture strength is almost unchanged for heat treatment at 1473 K up to 12 h, reaches a maximum of 1493 MPa at 24 h (which is consistent with the Vickers hardness results), and then remains almost constant with increasing aging time. This indicates that for aging periods longer than 24 h, the precipitation and growth of NbCr₂ has little effect on the fracture strength. However, the crack initiation strength decreases up to 12 h of aging, then increases up to 48 h, and decreases significantly by 72 h. The volume fraction of the NbCr₂ Laves phase increases with increasing aging time, and has a significant impact on the crack initiation strength. The volume fraction of the Laves phase is around 10% for the alloy aged for 48 h. In the early stages of microstructural evolution, when there is a lower volume fraction of the Laves phase, the crack initiation strength is expected to be governed by the BCC₁/BCC₂ structure. When the volume fraction of the Laves phase exceeds a critical value, cracking can then occur easily. Therefore, a critical volume fraction of the Laves phase may exist where the highest crack initiation strength can be obtained.

Fig. 11

Effect of heat treatment time at 1473 K on strength of 50Cr–30Mo–20Nb alloy.

4. Conclusions

A three-phase alloy with a composition of 50Cr–30Mo–20Nb was studied. The microstructural evolution and precipitation of the Laves phase were investigated using SEM, SAED, and TEM, and the fracture behavior was examined using Vickers hardness and compression tests at room temperature. The results are summarized as follows:

1. (1)    During aging at 1473 K after solution treatment, fast intragranular precipitation of the Cr-rich BCC₁ phase induces precipitation of alternating BCC₁/BCC₂ phases through a DP process, and this morphology remains stable with increasing aging time.
2. (2)    The Laves phase precipitates at BCC phase GBs and BCC₁/BCC₂ interphase boundaries. The growth of the Laves phase in BCC₁/BCC₂ two-phase areas consumes the Cr-rich BCC₁ phase in the vicinity.
3. (3)    Higher-temperature heat treatment results in higher BCC decomposition and Laves phase precipitation rates, while at 1773 K no BCC decomposition occurs and a BCC/Laves two-phase structure is formed.
4. (4)    The alloy aged at 1473 K for 72 h following solution treatment at 1973 K exhibits a hardness of 839 ± 8 HV and has better toughness than that for an alloy directly annealed at 1473 K because no obvious microcracks were observed after the Vickers hardness test, even under a load of 0.5 kgf for 30 s.
5. (5)    A minimum hardness and a highest fracture strength of 1493 MPa are obtained for the alloy aged at 1473 K for 24 h, where the two BCC phases dominate the microstructure. The volume fraction of the NbCr₂ Laves phase has a relatively small effect on the crack initiation strength.

Acknowledgments

This work was supported by a grant from the Advanced Low Carbon Technology Research and Development Program (ALCA) of the Japan Science and Technology Agency (JST) (No. JPMJAL1407). A part of this work was conducted at the Laboratory of Nano-Micro Materials Analysis, Hokkaido University, supported by the “Nanotechnology Platform” Program of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

REFERENCES

• 1)   B.P.  Bewlay,  M.R.  Jackson,  J.C.  Zhao and  P.R.  Subramanian: Metall. Mater. Trans. A 34 (2003) 2043–2052.
• 2)   K.  Li,  S.  Li,  Y.  Xue and  H.  Fu: Int. J. Refract. Met. Hard Mater. 36 (2013) 154–161.
• 3)   D.J.  Thoma and  J.H.  Perepezko: J. Alloys Compd. 224 (1995) 330–341.
• 4)   T.  Yamanouchi and  S.  Miura: Mater. Trans. 59 (2018) 546–555.
• 5)   K.  Li,  R.  Hao,  C.  Di and  D.  Gong: J. Mater. Res. 31 (2016) 2214–2222.
• 6)   A.J.  Knowles,  A.  Bhowmik,  S.  Purkayastha,  N.G.  Jones,  F.  Giuliani,  W.J.  Clegg,  D.  Dye and  H.J.  Stone: Scr. Mater. 140 (2017) 59–62.
• 7)   M.  Yoshida and  T.  Takasugi: Mater. Sci. Eng. A 224 (1997) 69–76.
• 8)   M.  Yoshida and  T.  Takasugi: Mater. Sci. Eng. A 224 (1997) 77–86.
• 9)   T.  Murakumo,  Y.  Koizumi,  K.  Kobayashi and  H.  Harada: Superalloys (2004) 155–162.
• 10)   T.M.  Pollock and  S.  Tin: J. Propuls. Power 22 (2006) 361–374.
• 11)   L.  Peng,  K.  Ikeda,  T.  Horiuchi and  S.  Miura: Mater. Trans. 59 (2018) 1706–1715.
• 12)   I.  Manna: Interface Sci. 6 (1998) 113–131.
• 13)   S.  Ueta,  M.  Hida and  M.  Kajihara: ISIJ Int. 53 (2013) 347–355.
• 14)   P.  Zięba: Arch. Metall. Mater. 62 (2017) 955–968.
• 15)   H.  Ramanarayan and  T.A.  Abinandanan: Physica A 318 (2003) 213–219.
• 16)   J.M.  Park: Appl. Math. Comput. 330 (2018) 11–22.
• 17)   L.Q.  Chen: Acta Metall. Mater. 42 (1994) 3503–3513.
• 18)   S.  Puri: J. Phys. Condens. Matter 17 (2005) R101–R142.
• 19)   K.C.  Chen,  S.M.  Allen and  J.D.  Livingston: J. Mater. Res. 12 (1997) 1472–1480.
• 20)   K.S.  Kumar and  C.T.  Liu: Acta Mater. 45 (1997) 3671–3686.
• 21)   H.R.  Cui,  F.  Sun,  K.  Chen,  L.T.  Zhang,  R.C.  Wan,  A.D.  Shan and  J.S.  Wu: Mater. Sci. Eng. A 527 (2010) 7505–7509.
• 22)   Y.  Xu,  M.  Wang,  Y.  Wang,  T.  Gu,  L.  Chen,  X.  Zhou,  Q.  Ma,  Y.  Liu and  J.  Huang: J. Alloys Compd. 621 (2015) 93–98.
• 23)   S.  Miura,  T.  Hatabata,  T.  Okawa and  T.  Mohri: Metall. Mater. Trans. A 45 (2014) 1136–1147.
• 24)   M.  Krüger,  P.  Jain,  K.S.  Kumar and  M.  Heilmaier: Intermetallics 48 (2014) 10–18.
• 25)   J.H.  Zhu,  C.T.  Liu and  P.K.  Liaw: Intermetallics 7 (1999) 1011–1016.
• 26)   D.  Casellas,  J.  Caro,  S.  Molas,  J.M.  Prado and  I.  Valls: Acta Mater. 55 (2007) 4277–4286.
• 27)   A.  Von Keitz and  G.  Sauthoff: Intermetallics 10 (2002) 497–510.
• 28)   D.J.  Thoma,  F.  Chu and  P.  Peralta: Mater. Sci. Eng. A 239–240 (1997) 251–259.
• 29)   J.A.  Hanna,  I.  Baker,  M.W.  Wittmann and  P.R.  Munroe: J. Mater. Res. 20 (2005) 791–795.
• 30)  H.J. Dorantes-Rosales, V.M. Lopez-Hirata, J.L. Gonzalez-Velazquez, N. Cayetano-Castro and M.L. Saucedo-Muñoz: Superalloys, (InTech, 2015) pp. 77–100.
• 31)  M.L. Saucedo-Muñoz and V.M. Lopez-Hirata: Scanning Electron Microscopy, (InTech, 2012).
• 32)   T.  Takasugi,  M.  Yoshida and  S.  Hanada: J. Mater. Res. 10 (1995) 2463–2470.
• 33)   T.  Takasugi,  M.  Yoshida and  S.  Hanada: Acta Mater. 44 (1996) 669–674.
• 34)   K.W.  Li,  S.M.  Li,  Y.L.  Xue and  H.Z.  Fu: Int. J. Refract. Met. H 36 (2013) 154–161.