2018 Volume 59 Issue 5 Pages 764-770
The refractory Mo–9Si–18B (at%) alloy doped with different mass fractions of nano-La2O3 were prepared by combining arc-melting, mechanical alloying and vacuum hot-pressing sintering techniques. The microstructures and mechanical properties of the alloys were systematically investigated. The results showed that the doped alloys presented finer grains than that of the non-doped alloys. Interestingly, most of the La2O3 particles dispersed in the grains interior rather than gathered in the grain boundaries by this combining process. This optimal microstructure had led to corresponding improvement in the compression strength (both at room temperature and 1200°C) and room temperature fracture toughness of the Mo–9Si–18B alloys. Among these alloys, the 0.6 mass% La2O3 doped alloys exhibited the best mechanical properties. The predominant strengthening mechanisms are particles dispersion strengthening and fine-grain strengthening. And the toughening mechanism mainly involves particle toughening.
In recent years, the research of advanced ultrahigh temperature structure materials have been principally driven by the needs to improve the efficiency of aerospace and powder-generation gas-turbine engines.1–3) Up to now, the best turbine blades made of Ni-based single-crystal super alloys can operate at the temperature of about 1150°C, i.e., close to 90% of their melting points.4,5) One way to further increasing in efficiency is to develop new ultrahigh temperature structural materials. Multiphase Mo–Si–B alloys, as the potential candidates of next generation ultrahigh temperature structural materials, have achieved considerable attention due to their high melting point, salient high temperature strength, excellent high temperature creep resistance and oxidation resistance, etc.6)
In the Mo–Si–B system, the two phase alloys that consist of intermetallic phases of Mo5SiB2 and α-Mo solid solution have received an increasing attention due to their potential applications for ultrahigh temperature hostile environments.7–9) The intermetallic phases of Mo5SiB2 in the alloys exhibit not only excellent high temperature strength and creep resistance but also outstanding oxidation resistance owning to the formation of borosilicate, which flow over the surface producing a protective scale.10–12) However, the Mo5SiB2 phase is very brittle at room temperature and consequently exhibits very poor toughness, with the fracture toughness value of only ∼2 MPa·m1/2.1) The presence of the ductile α-Mo phases, which have the fracture toughness value of ∼10 MPa·m1/2 at room temperature and ∼20 MPa·m1/2 at 1200°C,13) can improve the fracture toughness of the two phase alloys notably. But the oxidation resistance and strength of the alloys at ultrahigh temperature will be reduced remarkably at the meantime. The specific properties are determined by the volume fractions of the two phases and the microstructural morphologies, etc.7,14–17) One strategy for the best oxidation resistance and strength of the two phase alloys at ultrahigh temperature is to reduce the volume fraction of α-Mo phases by improving the fracture toughness of the α-Mo phases. In this way, the same fracture toughness of the alloys can be afforded by less α-Mo, thus one can expect the improved oxidation resistance and strength by increasing the content of the Mo5SiB2.1)
Up to now, numerous methods have been employed to improve the toughness of α-Mo phases, including micro-alloying with zirconium,18,19) MgAl2O4 spinel20) and Y2O3.21) In Liu et al.’s work, both strength and fracture toughness of the La2O3 doped Mo alloys have been improved remarkably by liquid-liquid doping technique, which led to an core-shell microstructure with nano-La2O3 particles uniformly distributed in the sub-micrometer grain interior.22) Correspondingly, we can expect that the strength and fracture toughness of Mo–Si–B alloys will be also significantly enhanced with most La2O3 particles uniformly dispersed in the grain interior, rather than concentrated in the grain boundaries. In this work, La2O3 doped Mo–9Si–18B (at%) alloys with an optimal microstructure, have been fabricated by a comprehensive processing combining arc-melting, mechanical alloying23) and vacuum hot-pressing sintering24) successfully. The drawbacks of composition segregation and inadequate reaction, resulted from the use of single technique to prepare Mo–Si–B alloys, could be validly avoided. And then the relationship between the mechanical properties of the prepared alloys and the mass fraction of doped La2O3 has been investigated. The strengthening and toughening mechanisms have also been discussed.
The alloy powders with nominal composition of Mo–9Si–18B (at%) doped with different mass fraction (0, 0.3 mass%, 0.6 mass%, 0.9 mass%) of La2O3 were prepared using high purity elements Mo, Si and B powders, with 99.95 mass%, 99.99 mass% and 99.95 mass% purity respectively. The four kinds of alloys, Mo–9Si–18B, Mo–9Si–18B+0.3 mass%La2O3, Mo–9Si–18B+0.6 mass%La2O3 and Mo–9Si–18B+0.9 mass%La2O3 are abbreviated as MSB0, MSB3, MSB6, MSB9 hereafter. Here the particle size of La2O3 was around 150 nm. In order to mix homogeneously, the powders were mixed in a planetary ball mill (Retsch® PM 100) in the argon atmosphere (0.5 MPa). To promote striking arc, the powders were pre-sintered into bulks with the pressure of 20 MPa in vacuum. Before each melting of Mo–Si–B alloys, a pure Ti bulk sample was melted to eliminate the residual O2 and N2 in the smelting furnace. After melted repeatedly for five times to obtain a chemical homogeneity in a partial pressure of argon atmosphere, the alloys were cooled in the water-cooled copper mold. Subsequently, the ingots were crushed into small bulks and milled in the same planetary ball mill in the argon atmosphere (0.5 MPa). The final average particle size after milling is around 4 µm. Finally, the powders were hot pressing sintered in vacuum at 1700°C with a pressure of 50 MPa for 2 hours.
The phases of alloys were determined by X-ray diffractometer (XRD, D/MAX-2400). Microstructural analysis was carried out using High Resolution Transmission Electron Microscopy (HRTEM, JEM-2100F) and Scanning Electron Microscopy (SEM, JSM-7000F) both equipped with Energy Dispersion Spectrometer (EDS). The final thinning and perforation of samples for HRTEM was accomplished by using ion beam thinner.
Uniaxial compression tests at room temperature and high temperature were carried out on the SUNS Electronic Universal Testing Machine (CMT5105) and the Numerical Control Dynamic Thermal-mechanics Simulation Testing Machine (Gleeble1500D) respectively at the same nominal strain rate of 5 × 10−4 s−1. The samples for uniaxial compression with a cross section of 2.5 mm × 2.5 mm and a height of 5 mm with the both ends of the specimens parallel to each other.
The fracture toughness values were determined by tree-point bending test. The polished specimens were 3 × 4 mm with a length of 26 mm and a notch tip depth of 2 mm. The tests were performed in SUNS Electronic Universal Testing Machine with a 20 mm span in room temperature with indenter speed rate of 10 µm·s−1. The details and specific parameters concerning the fracture toughness measurement are available in Ref. 25). In this paper, all the experimental results were the average values of three samples.
Figure 1 shows the XRD patterns of MSB0, MSB3, MSB6 and MSB9 alloys. All the alloys are composed of the same two phases of α-Mo and Mo5SiB2, which is complied with the schematic section of the 1600°C isothermal ternary Mo–Si–B phase diagram by Nunes et al.26) Further analysis reveals that the alloys contain approximately 28 vol% of α-Mo phases and 72 vol% of Mo5SiB2 phases respectively.
XRD patterns of Mo–9Si–18B alloys doped with different mass fraction of La2O3.
Representative back-scattered electron micrographs of MSB0, MSB3, MSB6 and MSB9 are shown in Fig. 2. Based on the XRD and the EDS analysis, the gray phases are Mo5SiB2 that have the highest volume fraction, and the bright phases are α-Mo. Besides the two phases, a few micrometers sized residual pores also exist, just as the black spots indicated. These pores may form during sintering, and they unlikely strongly influenced the mechanical properties since the sizes of the pores are small.27)
Back-scattered electron micrographs of alloys: (a) MSB0, (b) MSB3, (c) MSB6, (d) MSB9.
According to the quantitative metallographic and statistical techniques, the mean grain sizes of the alloys have been determined and listed in Table 1. We can see that both the grain sizes of α-Mo and Mo5SiB2 phases are reduced with the increasing addition of La2O3. According to Zhang’s work,28) the first reason may be that parts of La2O3 act nucleation agents and enhance the nucleation density. The other reason is the La2O3 particles will hinder the growth of the grains via pinning effect. As a result, the higher quantities of the addition of La2O3, the finer grain sizes will be obtained.
Figure 3 shows the TEM images of MSB0, MSB3, MSB6 and MSB9 alloys. It is clear that the alloys are composed by α-Mo and Mo5SiB2 phases. In addition, La2O3 particles almost homogeneously dispersed in both α-Mo and Mo5SiB2 grains in the MSB3, MSB6 and MSB9 alloys (Fig. 3(c)–(h)). Figure 4 clearly presents the details of morphologies and specific distributions of La2O3 particles in the (a) MSB0, (b) MSB3, (c) MSB6 and (d) MSB9 with high magnification, respectively. We can clearly found that particles with diameters of several tens nanometers appear in the interior of the grains. Further analysis by using EDS proves that these particles are La2O3, just as Fig. 5 indicated. By comparing, we found that the distribution of the La2O3 particles, effecting on the strength and toughness of the alloys, directly connected with its mass fraction. As the mass fraction increases, the quantities of La2O3 particles dispersed in the grains increase, and the mean particles sizes of La2O3 particles enhance simultaneously, which are listed in the Table 1. But the particles tend to cluster when the addition of the La2O3 reaches to 0.9 mass% (Fig. 4(d)). This might deteriorate the mechanical properties of the alloys.
TEM images showing the microstructure of La2O3 doped Mo–9Si–18B alloys. (a) is the bright field image of MSB0, (b) shows the selected area diffraction pattern of the indicated regions, (c)–(d), (e)–(f) and (g)–(h) are the corresponding bright and dark field images showing microstructure and the La2O3 particles distribution in the same region of MSB3, MSB6 and MSB9, respectively.
TEM images showing the details of morphologies and specific distributions of La2O3 particles in the grains, (a) MSB0, (b) MSB3, (c) MSB6 and (d) MSB9. The La2O3 particles were indicated by the arrows.
The Energy Dispersion Spectrometer of the La2O3 particles in the grains of MSB9.
The fracture toughness of alloys at room temperature is experimentally determined and the results are listed in Table 2. On one hand, the fracture toughness of doped alloys was higher than that of the non-doped alloys. This shows that the fracture toughness can be improved by doping La2O3 in the alloys. On the other hand, the fracture toughness of MSB6 is the highest among the alloys, and is better than the MSB9.
The typical compressive stress-strain curves of alloys at room temperature are shown in Fig. 6(a). It is obvious that the compressive strength of the doped alloys is higher than that of the non-doped alloys. Moreover, the compressive strength increases with the increasing of La2O3 till up to 0.6 mass%. Above 0.6 mass%, the compressive strength began to decrease. Thus, the MSB6 alloys have the highest compressive strength of 3294 MPa. Similar to compression strength, the strains have a similar variation tendency. By comparison, the alloys have higher yield strength at room temperature with more addition of La2O3 up to 0.9 mass% and the highest value of the yield strength is 2147 MPa in MSB9.
Compression stress-strain curves of Mo–9Si–18B alloys doped with different mass fraction of La2O3: (a) at room temperature, (b) at 1200°C.
The compression behaviors at 1200°C, which presented in Fig. 6(b), were different from that at room temperature. All the specimens were deformed to the strains of 13%. It is clear that the alloys display compressive plasticity at 1200°C with a peak flow stress followed by flow softening, which is consistent with Alur et al.’s work.29) Furthermore, both the peak flow stress and yield strength are highest when the doped La2O3 is 0.6 mass% at 1200°C. The detailed mechanical properties of the alloys were listed in Table 2.
In room temperature, toughening in Mo–Si–B alloys can be considered as two different classes of mechanisms, intrinsic and extrinsic toughening mechanisms.29) The performance of intrinsic toughening, referred to crack trapping, will be enhanced with a higher volume fraction of α-Mo. The extrinsic toughening, afforded by the uncracked ligament bridging, is more potent for coarser-grained alloys.1) As mentioned above, the MSB0, MSB3, MSB6 and MSB9 alloys contain the same amount (around 28 vol%) of α-Mo. So the crack trapping have a negligible effect on improving the toughness of the doped alloys. However, in terms of extrinsic toughening mechanisms, the addition of La2O3 is detrimental to the toughness because it will reduce the grain size of α-Mo.
In this work, another main toughening mechanism is particle toughening, of which the detailed mechanisms are micro-cracking, crack deflection and so on.28) The fracture surfaces of each sample were present in Fig. 7. The intergranular fracture is the mainly fracture pattern in the MSB0 alloy (Fig. 7(a)). The fracture pattern gradually changed from mainly intergranular fracture to mainly transgranular fracture with the increase of La2O3 particles (Fig. 7(b)–(d)). A large number of secondary cracks and the uneven rupture surface appear in the MSB6 alloy (Fig. 7(c)) due to the uniform distribution and moderate number of La2O3 particle. These illustrate that many crack deflections occur in the process of main crack extension, which are beneficial to the improvement of the fracture toughness, because more energy will be absorbed during the process of cracks propagation. As the addition of La2O3 increases, the distances between La2O3 particles decrease, even some particles cluster somewhere, as shown in the Fig. 4(d). These could induce stress concentration in interface between the substrate and La2O3 clusters, which is unfavorable to crack deflection and made the rupture surface flat without secondary cracks (Fig. 7(d)). Correspondingly, the fracture toughness decreases in MSB9 alloy.
Fracture surface morphology of the La2O3 doped Mo–9Si–18B alloys by the fracture toughness test: (a) MSB0, (b) MSB3, (c) MSB6 and (d) MSB9.
The microstructure analyses show that the contents of doped La2O3 have notable effect on grain sizes of alloys, distances between La2O3 particles and the sizes of La2O3 particles. More addition of La2O3 results in finer grain sizes, shorter distance and coarser La2O3 particles. Moreover, the La2O3 particles distributed in grains will hinder the dislocation movement. Thus, the strengthen mechanisms of the La2O3 doped alloys should include fine-grain strengthening and particles dispersion strengthening.
The researched Mo–9Si–18B alloys contain α-Mo phases and the intermetallic phases of Mo5SiB2. At room temperature, the α-Mo phase is the mainly ductile phase, as few mobile dislocations and hence little plasticity can be realized in the highly ordered Mo5SiB2 phase.1,30) Considering the little plasticity and high fracture strength of the Mo5SiB2 phases, one can draw a conclusion that the yield phenomenon mainly occurred in the α-Mo phase and the Mo5SiB2 phase only show elastic deformation before fracture. Thus, for the convenience of calculation, we suppose that the effect of fine-grain strengthening and particles dispersion strengthening are only related with the α-Mo phase.
Therefore, the yield strength of the alloys can be quantitatively calculated by superimposition of strength of material with infinite grain size, σ0, fine-grain strengthening term, $\text{K}/D^{\frac{1}{2}}$, and particle strengthening term, σOR, which is described as follows:
| \begin{equation} \sigma_{y} = \sigma_{0} + K/D^{\frac{1}{2}} + \sigma_{OR} \end{equation} | (1) |
As shown in Fig. 4, the sizes of La2O3 particles distributed in grains and the distances between each other change with the mass fraction of doped La2O3. According to Orowan-Ashiby equation, the yield strength increased by La2O3 particles strengthening can be determined as follows:31–34)
| \begin{equation} \sigma_{\textit{OR}} = \cfrac{MGb}{(1.18)\cdot 2\pi\cdot\varphi\cdot\Biggl(\sqrt{\cfrac{\pi}{6f}} - 1\Biggr)}\ln\left(\frac{\varphi}{2b}\right) \end{equation} | (2) |
According to the Formula 1, we expect the experimental value σy-σOR meet the Hall-Petch relationship, i.e., the value has linear relation to the reciprocal square root of the grain size of α-Mo phases, with an intercept value of matrix strength, σ0. The relationship is represented in Fig. 8. And we conclude that the experimental data are close to a line with slope of 2230 MPa•µm1/2 and intercept of 649 MPa. Thus Formula 1 can be written as:
| \begin{equation} \sigma_{y} = 649 + \frac{2230}{D^{\frac{1}{2}}} + \sigma_{OR} \end{equation} | (3) |
Hall-Petch representations of yield stress after removing the Orowan strength.
As is shown in Fig. 4(d), the distances between La2O3 particles decline. Even some particles tend to cluster when the mass fraction of doped La2O3 is 0.9 mass%. This situation is harmful to the plasticity and fracture toughness of the alloys, because the distances between the La2O3 particles are too small to be bypassed by the dislocations, hence lead to pile-up of the dislocations. The pile-up of the dislocations will give rise to stress concentration easily around the particles, which is inclined to generate cracks and results in the failure of the alloys. Thus the compression strength and plasticity will decrease when the mass fraction of the doped La2O3 exceed to the optimal quantity, i.e. 0.6 mass% in this work. That is why the compression strength and plasticity of MSB9 are lower than that of the MSB6.
As a result, the moderate addition of La2O3 to Mo–9Si–18B alloys will increase the compression strength and plasticity of the alloys, while excess additions will worsen the mechanical properties at room temperature.
4.3 Strengthening mechanism at high temperatureAt high temperature, the movements of dislocations become easier and more dislocations will be activated due to thermal activation. So the strength of alloys decreases, however, the plasticity increases. Besides of the fine-grain strengthening and particle strengthening mechanism, the grain boundary sliding also affects mechanical properties of alloys potently.30,36) The finer grain size alloys will be easer subjected to grain boundary sliding at high temperature due to more amounts of grain boundaries.
Table 1 clearly shows that the addition of La2O3 can refine the grain size of the Mo–9Si–18B alloys, i.e., increase the amounts of grain boundaries of the alloys. Thus, at high temperature, the addition of La2O3 to Mo–9Si–18B alloys will increase the strength of the alloys through the fine-grain strengthening and particle strengthening mechanism.37) In addition, it will worsen the strength of alloys by the grain boundary sliding.38) So when the addition of La2O3 to Mo–9Si–18B alloys exceed to the optimal point, the strength of alloys will decrease, just as the Fig. 6(b) shown.
Some studies have shown that the Mo alloys doped with La particles and Al2O3 particles have a higher oxide resistance.39,40) Similarly, the microstructure with the La2O3 particles dispersed in the interior of the grains should possess a good oxidation resistance at high temperature, and this property requires further investigation.
The Mo–9Si–18B (at%) alloys doped with different mass fraction of La2O3 were fabricated using a comprehensive route by combining arc-melting, mechanical alloying and vacuum hot-pressing sintering techniques. Then the microstructure and mechanical properties of the alloys were researched for comparison. Thus the following conclusions can be drawn from this study:
The financial support from the National Natural Science Foundation of China (NSFC) under Grant no. 51621063 was gratefully acknowledged. This work was also supported by the foundation of National Ministry and Commission (granted no. 613262).
