Powder production, FAST processing and properties of a Nb-silicide based alloy for high temperature aerospace applications
Release time:
2025-09-25
1. Introduction
To improve the performance of gas turbine engines the turbine entry temperature (TET) must be increased to around 1850 ◦C. Efficiency can also be improved by reducing the overall weight of components (increasing the thrust to weight ratio). New metallic ultra-high temperature materials (UHTMs) are required that (i) have lower densities than conventional Ni-based superalloys, (ii) can withstand higher operating temperatures and (iii) meet property goals for toughness, creep and oxidation resistance [1]. Candidate metallic UHTMs are refractory metal (RM) intermetallic composites (RMICs), refractory metal high entropy alloys (RHEAs) and refractory metal complex concentrated alloys (RCCAs) [2].
RMICs based on Nb, i.e., RM(Nb)ICs, also known as Nb-silicide based alloys or Nb-silicide in situ composites, are composed of Nb solid solution (Nbss) and intermetallics (e.g. Nb5Si3, Nb3Si, C14–NbCr2 Laves phase and A15 compounds, e.g., Nb3Sn, Nb3Al, Nb3Si), currently are explored as next generation metallic UHTMs to be used in the hottest parts of a turbine engine [1]. Some of the RM(Nb)ICs also satisfy (comply with) the definition of RCCAs or RHEAs, i.e., they are RM(Nb) ICs/RCCAs or RM(Nb)ICs/RHEAs. These alloys can have high melting temperatures in excess of 1750 ◦C, excellent high temperature mechanical properties, and lower densities (e.g. between 6.5 and 7.9 g/cm3) than Ni-based superalloys [1,3]. The room temperature fracture toughness and oxidation resistance of Nb-silicide based alloys has been improved through alloy design and the use of specific alloying elements [1]. Like the Ni-based superalloys, the metallic UHTMs will need the protection offered by environmental coatings [4]. Unlike the Ni-based superalloys, the microstructures of metallic UHTMs are susceptible to interstitial contamination [5], in particular the bcc solid solution [6].Interstitial contamination, which can have a significant effect on the properties of RMICs, RHEAs and RCCAs, imposes constraints on alloy design and processing [1].
Manufacture and upscale of Nb-silicide based alloys is a known challenge for these materials [1]. Alloys can be produced using conventional casting methods using cold hearths, for example arc-melting or plasma melting with water-cooled Cu crucibles or directional solidification [1,7,8]. However, these methods result in compositional variation across the ingot, macrosegregation of solute additions [9], heterogeneous microstructure and casting defects such as cracks and pores [7]. Advanced manufacturing methods have been proposed to overcome these issues and allow greater control over the microstructure. More recently, powder metallurgy (PM) techniques, including Field Assisted Sintering Technology (FAST), also known as Spark Plasma Sintering (SPS), have been used to produce small amounts of Nb-silicide based alloys, however the volume of published research is relatively limited. Furthermore, interstitial contamination in FAST processing of Nb-silicide based alloys has not received the focus it deserves, as is the general case for the processing of these alloys [1].
FAST is an effective sintering method for many different powdered materials, including metals and alloys [8,10]. It is capable of consolidating to very high densities whilst requiring relatively low temperatures and processing times compared with Hot Isostatic Pressing (HIP). This is due to simultaneous application of heat and pressure, along with rapid heating rates. FAST is therefore a suitable technique for the production of Nb-silicide based alloys from powders and has been shown to give fully dense material whilst eliminating casting defects and generating a more refined microstructure [7,11,12].
Several different approaches to producing Nb-silicide based alloys using FAST have been investigated in previous studies [7,11–13]. Blends of the necessary elemental powders have been used to generate alloys via reactive sintering, although this is difficult to control, due to the differing melting temperatures in multicomponent systems, especially in those with the addition of other RMs such as Mo, Ta or W, and sluggish reaction kinetics [14]. Instead, for binary Nb–Si alloys mixtures of Nb and Nb5Si3 powders have been used to overcome the lower melting point of Si. However, for multicomponent alloy systems with RMs and other low melting point elements, e.g., Al and Sn, the use of pre-alloyed powders is preferred [3,7]. Pre-alloyed powders can be produced using different manufacturing techniques. Mechanical alloying or comminution of either arc melted ingots or pre-sintered elemental powders via ball milling have been used on smaller scales [7,11], whereas inert gas atomisation has been used at larger scales [3,12,13,15].
The limited published literature on FAST of Nb–Si alloys relates to manufacture of samples typically less than 20 mm in diameter [7,11,12]. The capabilities of FAST are continually increasing, with the largest machines now capable of producing parts up to 600 mm in diameter, highlighting that further scale up is already possible [16]. The refined and more homogeneous microstructures generated via this powder-based route are expected to provide improved and more consistent properties than from casting alone, whilst eliminating defects. This is crucial for materials in critical components, where properties such as fatigue performance are dependent on a homogeneous, defect free microstructure. Larger scale FAST will allow for more representative mechanical property studies to be conducted, for example to assess tensile and compressive properties at elevated temperatures using electro-thermal mechanical testing (ETMT). Furthermore, data about interstitial contamination of PM Nb–Si alloys is lacking even though interstitials are known to affect phase stability and mechanical properties [1,6,17].
In this study, samples up to 60 mm in diameter of Nb-silicide based alloy with nominal composition Nb–18Ti–22Si–6Mo-1.5Cr–2Sn-1Hf (at. %) were produced through a processing route of arc melting, ballmilling, and FAST consolidation. The composition was chosen to give a good balance of high temperature mechanical properties and oxidation resistance. Cr, Hf, Mo, Sn, Ti partition to the Nbss and provide solid solution strengthening, in particular Mo and Ti, whilst Ti and Hf partition to the Nb5Si3. Ti also improves oxidation resistance, reduces density, and improves fracture toughness [8,18]. Si was added to form the desirable Nb5Si3 silicide. This can be a tetragonal phase (tI32) which can have both a high temperature stable βNb5Si3 phase (prototype W5Si3) and low temperature stable αNb5Si3 phase (prototype Cr5B3) [17]. Metastable γNb5Si3 (hexagonal hP16, Mn5Si3 prototype) can be stabilised by carbon contamination [17] and by high Ti and Hf concentrations in the silicide, resulting in Nb/(Ti + Hf) < 1 [18,19]. Furthermore, it was expected that Cr, Sn and Hf additions would improve the oxidation resistance, in particular Hf which scavenges oxygen [18] and decreases the diffusivity of oxygen towards the bulk of Nb-silicide based alloys and thus is expected to be effective regarding control of contamination with oxygen [20].
This paper presents the results of an initial study of rocessmicrostructure relationships for the aforementioned alloy, as well as assessing mechanical properties through hardness measurements and electro-thermal mechanical testing (ETMT).
2. Methods
High purity pieces of Nb (99.8 %), Ti (99.99 %), Si (99.999 %), Mo (99.95 %), Cr (99.99 %), Sn (99.99 %) and Hf (99.9 %) were weighed out based on the alloy composition. An Arcast Arc 200 arc melter with non-consumable tungsten electrode was used to melt the elements together in a water cooled copper crucible. The melting chamber was pumped down to a vacuum of 5 × 10− 2 mbar and flushed with argon two times before achieving the ultimate vacuum level of 4 × 10− 5 mbar. Then, the chamber was backfilled with argon to about 700 mbar before melting. Arc melting was done in two stages to avoid severe vaporisation of elements with high vapour pressure, with the ingots being flipped and re-melted six times in each stage to ensure best possible chemical homogeneity of the alloy. In the first stage, the elements with higher melting point (Nb, Ti, Mo, Cr and Hf) were melted together, then Si and Sn were added in the second stage. Extra Si, Sn and Cr were added to compensate for their loss due to vaporisation during arc melting. Several button ingots were produced for further processing, each approximately 100 g in mass and around 45 mm in diameter. Different stages of the processing of the Nb-silicide based alloy of this study are shown in Fig. 1.
The ingots were initially crushed using a percussion mortar. The crushed particulate material was then placed into a zirconia-lined jar, along with zirconia grinding balls, fitted with an aeration lid and flushed with argon to create an inert atmosphere. A Retsch PM100 planetary ball mill was used to grind the material to a powder under the inert conditions. Individual ingot crushed and ground powders were combined to create a single powder blend. The powder was sieved to <250 μm and the resulting particle size distribution (PSD) was measured by laser diffraction using a Malvern Mastersizer 3000.
Initial FAST processing was carried out using 10 g of the alloy powder blend to produce 20 mm diameter samples. The FAST parameters that were optimised at the small scale subsequently were used to produce a sample of 60 mm diameter and 4.5 mm thickness, using 90 g of the powder. The powder was placed into a 60 mm internal diameter graphite mould, lined with graphite foil that was removed after processing. A graphite felt jacket was used around the tooling for thermal insulation to prevent heat loss and improve power efficiency. FAST processing was done using an FCT Systeme HP D 25 type furnace, where the temperature was measured using a pyrometer. The sample was heated at 150 ◦C min− 1 up to 1300 ◦C, then at 75 ◦C min− 1 to a maximum temperature of 1500 ◦C, at which point a pressure of 50 MPa was applied and held for 5 min. The entire test was performed under a vacuum of 5 × 10− 2 mbar. After cooling to room temperature, the sample was removed from the graphite tooling, and the surface was cleaned using abrasive grit blasting.
The as cast ingot, milled powder and FAST consolidated material were studied to understand the microstructural evolution and anychanges in chemistry during processing. The bulk chemistries of the powder blend and FAST samples were measured using inductively coupled plasma - optical emission spectroscopy (ICP-OES, for alloying elements), inert gas fusion (oxygen and nitrogen, Eltra ONH 2000), and combustion analysis (carbon). Measurement error is approximately ±10 at.%. Phases in the bulk as cast ingot and FAST sample were identified using a Bruker D2 phaser x-ray diffraction (XRD) machine (Cu radiation (30 kV, 10 mA), 0.02◦ step) and ICDD database. One of the ingots and a section of the FAST sample were sectioned using a diamond cutting wheel, then polished by metallographic preparation consisting of a range of SiC grinding papers, followed by 0.06 μm colloidal silica solution. An FEI Inspect F50 scanning electron microscope (SEM) and Oxford Instruments energy dispersive X-ray spectroscopy (X-EDS) detector were used to analyse theicrostructure. Vickers hardness was measured on the polished surfaces using a Struers Durascan-80. A grid of 16 indents covering an area of 8 mm2 was tested, with a force of 1 kg (HV 1) applied for 15 s for each indent.
Electrical discharge machining (EDM) was used to extract 3 x 1 × 50 mm matchstick specimens from the FAST consolidated material for ETMT. The density of the alloy was measured using the Archimedes method, with a mean average value taken from two of the matchstick specimens.
An ETMT system developed at the National Physical Laboratory was used to measure the macroscopic stress, while the plastic strain is calculated by monitoring changes in resistance. More details of the experimental set-up can be obtained from Refs. [21–25]. This ETMT system also is used to assess the mechanical behaviour of Ni-based superalloys [24]. The true plastic strain, εpl, was measured according to the change in resistance between the central gauge length of the test piece where the temperature profile is approximately uniform ( ± 5 K). The measurement resolution is determined by the type of thermocouple used t.o measure temperature, in this case Pt/Pt–13%Rh and is less than 1 K at 1573 K (1300 ◦C). In Ni-base superalloys, the individual resistivity of the γ and γ′ phases differ at any temperature (resistivity of γ′ being greater than that for γ phase) and therefore the temperature dependence of resistivity arises from the relative volume fraction of γ and γ′, irrespective of the application of load [22]. Since dissolution of γ′ occurs during heating, there will be a concomitant change in resistivity (and resistance) with temperature arising from the phase transformation. Consequently, in the case of applied load, R is the measured resistance and R0 is the strain-free resistance for zero-load, equation (1) is used to calculate the plastic strain, ϵpl [22,23]:
In miniature testing, the gauge volume can be used to assess location specific properties. The gauge length here is determined by the distance between which resistance is measured; approximately L = 2.75 mm at room temperature for Ni-alloys [26–28]. During heating, the sample gauge length will increase, though this is small and therefore no correction is needed. Indeed, in the case of Nb-silicide based alloys, the strain increase ε = α ΔT ≪ 1, where the thermal expansion coefficient α = 7.3 × 10− 6 K− 1 and ΔT ≈ 1300 K. The gauge length at any temperature is L ~2.75 mm and therefore constant over the entire temperature range. The gauge volume is therefore typically about 5.5–11 mm3.
The variation of resistivity with temperature, ρT can be calculated via;
where RT is the ETMT measured resistance at temperature T, and A is the cross-section area.
Since the plastic strain is measured from the resistance, any error in the resistance measurement contributes to an error in the plastic strain. The current and voltage measurements are made with better than 1 part in 104 uncertainty and the maximum range in the measured resistance is ±0.002 mΩ and with an accompanying standard deviation of ±0.001 mΩ. From Equation (1), the upper and lower plastic bounds of plastic strain is given by; εpl = 12 ln ( R±ΔRR0±ΔR0). This is used to estimate a plastic strain resolution of about 0.05 %.
Compressive creep tests were carried out on matchstick specimens. The length of each specimen was 50 mm, while the gauge length (isothermal to within ± 3K between 1000 ◦C and 1200 ◦C was between 2.7 and 3 mm). The width and thickness of the specimens varied between 1.27 to 3 mm and 0.79–0.98 mm, respectively. The sample was heated under zero-load to the set temperature and a number of interrupted creep tests were carried out by stepping the load. The primary purpose for carrying out such stepped creep tests was to be able to obtain creep data over a range of stresses at a given temperature and thus obtain a steady-state creep rate, as this approach is sample economic. A similar approach has been adopted before [26]. To identify any history dependence of prior deformation (at the previous stress/temperature)on the creep rate at the next successive stress/temperature, a specimen was first heated to 1050 ◦C and a set of creep tests carried out, before heating to 1200 ◦C and carrying out some more creep tests. This will be referred to in the next section as the first creep test. The reverse was also done, i.e., another specimen was first heated to 1200 ◦C and a set of creep tests carried out, before cooling to 1050 ◦C and carrying out further creep tests.
3. Results and discussion
3.1. Blended Nb-silicide based alloy powder
Table 1 shows that there was good agreement between the nominal composition, ingot composition, and that of the powder blend. During the initial arc melting trials, vaporisation led to significant variation in the composition, and so a few iterations were completed using different amounts of some elements to achieve a chemical composition as close as
possible to the nominal one. Despite this, the actual composition of the ingots did not match exactly the nominal alloy composition. This accounts for the variation between the ingot compositions and powder blend, as the powder blend was the average of several combined ingots.
Interstitial elements can impact the mechanical properties of RMs, alloys with RM additions [1] and their phases [6]. The maximum solid solubility of oxygen in Nb is 9 at.% at 1915 ◦C and about 6 at.% at 1500 ◦C [29]. The effect of solute elements in the solid solutions in Nb silicide based alloys on the maximum solid solubility of oxygen is not known [6], and the same is the case for the Nbss of the alloy in this study. Contamination with oxygen of the bcc solid solution affects its hardness and Young’s modulus [6]. The Nb5Si3 can also be contaminated with oxygen, however its contamination is less severe compared with the Nbss [30] and affects the hardness and Young’s modulus of the silicide. The bcc solid solution in RM(Nb)ICs is their Achilles’ heel as it provides the path for interstitials to contaminate their microstructure [6]. Control of interstitials during processing is challenging as contamination can occur from several sources, including milling media and container, milling environment and tooling [1]. Interstitial pickup occurred during the processing, with a significant increase in oxygen content in the powder blend compared with the ingot, revealing that the majority of the increase in oxygen content occurred during the crushing and ball milling of the material. The oxygen content given in Table 1 is a bulk measurement, and therefore the oxygen content of individual phases is not reported. The increase in bulk oxygen content could have been caused by contamination from the ZrO2 milling media, although it is more likely from reaction with atmospheric oxygen. This suggests that the argon flushing of the milling container was not sufficient, however more investigation is required to confirm this. Heat generated during milling could also enhance pickup of oxygen. The increase in oxygen content can affect mechanical properties and change phase equilibria where contamination is high [31]. Therefore, it is important to both minimise and quantify the extent of contamination during processing, to assess the impact on properties.
The particle size distribution (PSD) of the powder was measured after milling and sieving, with the results displayed in Fig. 2. Although sieved to <250 μm, the D90 was <100 μm, indicating that the milling process was effective in producing a fine powder from the ingots, but some larger particles remained. With greater understanding of the milling process, the PSD could be tailored to achieve a specific PSD.
SEM analysis of the blended powder (Fig. 3a) showed an expected large variation in both the particle size and morphology. Particles exhibited irregular morphology resulting from the milling, with fine particles adhering to the surfaces of the larger particles, a feature that is common also with gas atomised particles where “satellite” fine powder particles adhere to larger particles. The backscattered electron micrograph of the polished cross section of powder particles in Fig. 3b reveals the microstructures and phases present: Nb5Si3 and Nbss (brighter Z contrast). The area fraction of each phase varied between different powder particles and also depended on powder particle size owing to the heterogeneous microstructure of the as cast ingots. There was no evidence of HfO2 particles in the microstructure of the as cast alloy and the powder particles after milling.
3.2. FAST consolidated Nb-silicide based alloy powder
Initial FAST parameters were chosen based on previous, small-scale, trials with similar alloy powder compositions [32]. Several 20 mm diameter samples were produced and studied to determine the optimal parameters required to achieve full density for the alloy of this study. It was determined that a temperature of 1500 ◦C, combined with a pressure of 50 MPa was sufficient to achieve full consolidation, with no remaining porosity visible. These conditions were again used but on a larger scale to create a 60 mm diameter sample (Fig. 1). This size would make it possible to extract specimens for the ETMT creep experiments.
After FAST consolidation, the content of interstitial elements was once again measured and compared with the powder, to determine any effect of the processing. There was no evidence of oxygen or N pickup (1.31 and 0.2 at.%, respectively), due to the FAST processing occurring under vacuum conditions. The C content can increase due to contamination from the graphite foil surrounding the sample during processing. In our experiments, the measured bulk C concentration increased from 0.04 to 0.09 at.% suggesting some pick up of C. It is expected that C will be concentrated at the surface of the sample, in contact with the graphite foil.
Bulk XRD data of the as cast and FAST material is given in Fig. 4. In the as cast alloy, both the αNb5Si3 and βNb5Si3 silicides were formed together with the bcc Nbss. In “large” as cast ingots of Nb-silicide based alloys the βNb5Si3 transforms to αNb5Si3 during solid state cooling [33,34]. In the cast alloy of this study the transformation of βNb5Si3 to αNb5Si3 must have started during solid state cooling, owing to the size of the ingot but was not finished owing to the Cr and Mo addition that tends to increase the stability of the βNb5Si3 (both Cr5Si3 and Mo5Si3 have the same prototype as βNb5Si3, namely W5Si3). Furthermore, hexagonal γNb5Si3 was not formed, which indicated that interstitial contamination of the microstructure was not severe and that the partitioning of Hf and Ti to Nb5Si3 did not result in Nb/(Ti + Hf) < 1 [17–19]. After FAST, the stable phases were bcc Nbss and the low temperature stable αNb5Si3 silicide. This suggests that the transformation of βNb5Si3 to αNb5Si3 was completed during FAST under these conditions. In addition, small peaks for HfO2 were identified in the FAST sample. This correlates with the increased oxygen content of the material after milling, as the Hf acts as an oxygen scavenger, forming HfO2. Furthermore, there was no evidence for γNb5Si3 in the FAST microstructure.
SEM and X-EDS were used to study the microstructure of the alloy after FAST processing, and to compare with the as cast ingot. The as cast microstructure (Fig. 5a and b) from the centre of the ingot, consisted of large Nb5Si3 regions surrounded by mixture of Nbss and Nb5Si3. There were some areas of eutectic consisting of these two phases, which were mostly formed adjacent to Nb5Si3 grains (seen in Fig. 5b). This would indicate that the solidification of the ingot started with the βNb5Si3 as the primary phase, followed by the formation of the Nbss and then of the Nbss+ βNb5Si3 eutectic. After milling and FAST consolidation, the microstructure changed considerably. The FAST microstructure appeared more homogeneous and refined, as expected from using powder. The large Nb5Si3 regions had been broken down during the milling process and so were much smaller in the FAST microstructure. There was still some variation in the scale and distribution of phases in the microstructure, as shown in Fig. 5d, resulting from the differences between the powder particles seen in Fig. 3b. Optimisation of the milling process and/or sieving to a smaller particle size would further improve microstructural homogenisation. Some porosity was observed, which is common in materials processed using PM, however this was measured to be <0.2 % following study of a mosaic of optical micrographs, and so not expected to have any significant impact on mechanical performance.
Elemental maps in X-EDS shown in Fig. 6 highlight how the different elements were segregated within the microstructure. As expected, Ti, Mo, Cr and Sn partitioned to the Nbss, whereas Si was present in the Nb5Si3 phase. A third phase is also seen in the FAST material, with X-EDS confirming this to be where the Hf is concentrated as HfO2.
3.3. Hardness and creep behaviour of the NbSi based alloy
The hardness of the alloy, both as cast and after FAST processing, was measured as described in section 2. This made it possible to compare this alloy with other Nb-silicide based alloys and also to highlight the effect of differences in microstructural inhomogeneity on the hardness. Although both the as cast and FAST material exhibited similar mean hardness values, respectively 764 and 816 HV (Table 2), these were higher than the calculated with the alloy design methodology NICE [18] hardness (746 HV) for the nominal alloy composition. The range of values for the as cast alloy was significantly greater due to its more heterogeneous microstructure. The hardness of the Nb5Si3 silicide is significantly higher than that of the Nbss. Thus, large regions of Nb5Si3 lead to measuring high hardness value. In contrast, the FAST microstructure is more refined and homogeneous, and the indents sample a more representative microstructure. The maximum and minimum alloy hardness values in Table 2 are within the range of hardness that is typical of RM(Nb)ICs and RM(Nb)ICs/RCCAs. The measured density of the alloy after FAST processing was 7.12 g/cm3. The density was lower than the density of latest generation Ni-based superalloys used in high pressure turbines in aero engines, in the range of values reported for RM (Nb)ICs [1] and significantly lower than the density of high strength RCCAs and RHEAs [35]. The room temperature yield strength of the alloy of this study calculated from hardness (σy = HV/3 – where hardness is in GPa) was 2497 and 2667 MPa, respectively for the as cast and FAST condition. For the latter, the yield strength was slightly higher than that of the RHEA nanocrystalline alloy MoNbTaVW (ρ = 12.29 g/cm3) processed using SPS [35]. The room temperature specific yield strength calculated from hardness and the measured density was 351 and 375 MPa cm3 g− 1, respectively for the as cast and FAST condition. For both conditions it was higher than that of the RHEAs and RCCAs reviewed in Ref. [35], in the range of values reported for B containing arc cast RM(Nb)ICs, RM(Nb)ICs/RCCAs [1,4], the Ge containing RM (Nb)IC/RCCA arc cast alloy ZF9 (Nb–24Ti–18Si–5Al–5Cr-5Ge-5Hf, at.%, nominal [36] and the RM(Nb)IC alloy ZF5 (Nb–24Ti–18Si–5Al-5Ge, at.%, nominal [37]), the RM(Nb)IC arc cast alloy JG3 (Nb–24Ti–18Si–5Al–5Cr–2Mo, at.%, nominal, [38]) and the arc cast alloy EZ5 (Nb–24Ti–18Si–5Al-5Hf-5Sn, at.%, nominal, [39]).
Fig. 7(a) and (b) plot the creep strain versus time corresponding to 1050 ◦C–1200 ◦C, respectively in case of the first creep test. Fig. 7(d) and (e) plot the creep strain versus time corresponding to 1200 ◦C–1050 ◦C, respectively and refer to the second creep test. In both cases it can be observed that steady-state prevails with a negligible occurrence of primary creep. The steady-state creep rates were calculated in Fig. 7(a) and (b) and Fig. 7(d) and (e) and plotted in a log-log plot in Fig. 7(c) and (f) respectively, from which the stress exponent (n) can be obtained. For creep tests performed at 1050 ◦C, n ~2.62 for the first creep test (1050 ◦C first, followed by 1200 ◦C) and n ~ 1.73 for the second creep test (1200 ◦C first, followed by 1050 ◦C). For creep tests performed at 1200 ◦C, n ~1.96 for the second creep test and n ~1.66 for the first creep test.
The stress exponent n of Nb5Si3 is about 1, higher for Nb and Nbss [18], between 2.5 and 3 for binary Nb–10Si and Nb–16Si alloys, in a wider range from about 1.1 to 11 for Nb–Si–Ti-Hf silicide based alloys [40], in the range 2–6 for MASC and MASC based alloys (the Nb–25Ti–16Si-8Hf-2Al–2Cr alloy (nominal, at.%) is known as MASC =metal and silicide composite) and in the range 1.3–4.5 for Nb-silicide based alloys with/without Ti, and RM (Mo,Ta,W) additions and in the range 4.35–6.35 for the single crystal superalloy CMSX4 [41]. The aforementioned is from compressive creep tests of phases and alloys that were creep tested in the as cast (arc melted or directionally solidified) or heat treated conditions and were not tested using ETMT.
The n values of the alloy of this work are close to those of binary Nb5Si3 and Nb–Si binary alloys discussed in Ref. [40], and some of the RM(Nb)ICs studied in the ULTMAT project [41]. The n values are also close to those of the NbCr2 Laves phase and A15-Nb3Al intermetallic [18], which can be present in the microstructures of RM(Nb)ICs and RM (Nb)Cs/RCCAs [1,13]. Differences in n values of RMICs are attributed to differences in chemical composition, microstructure or both. For example, the microstructures of the above mentioned binary Nb–Si alloys consist of large primary Nbss grains surrounded by Nbss + Nb5Si3 eutectic, the microstructures of Nb–Si–Ti–Al–Cr-Hf silicide based alloys comprise continuous or near continuous Nbss with dispersed silicides and Laves phase or both depending on alloy chemical composition. Furthermore, the n value of RM(Nb)ICs changes with vol% silicide [40].
The alloy of this study does not meet the creep property goal (i.e., the creep strength should be greater than 170 MPa at a creep rate of 2 ×10− 8 s− 1 at 1200 ◦C). This goal assumes ρ = 7 g/cm3 [1]) as its creep rate at 1200 ◦C/170 MPa was 1.75 10− 4 s− 1. Also, its creep rate was higher than that of the Ti containing RM(Nb)ICs discussed in Ref. [1], as would be expected owing to its higher Ti/Si ratio (0.87). At 1200 ◦C/100 MPa the creep rate (1.6 10− 6 s− 1) of the FAST alloy calculated with NICE [18] from the contributions made to creep by intrinsic resistances to dislocation motion [1,18] was similar to the experimental creep rate of the MASC alloy [40] but the value from the ETMT test (6.7 10− 5 s− 1) was one order higher. For the creep at 1050 ◦C/200 MPa the alloy methodology NICE indicated a key role for the parameter δ that depends on atomic size, as the calculated (for intrinsic resistance to dislocation motion) and measured creep rates were the same (7.35 10− 6 s− 1). The compressive creep test of single crystal CMSX4 at 1093 ◦C/200 MPa was 1.1 10− 6 s− 1 [41].
4. Conclusions
This study has demonstrated the use of FAST as part of a production route for Nb-silicide based alloys that are under development for applications in future high temperature turbine engines, and demonstrated that larger scale FAST processing than seen previously in the literature (i) can create fully dense, 60 mm diameter samples, (ii) brings new opportunities for scaling up production and (iii) allows for more extensive assessment of mechanical properties. In this study the compressive creep of a RMIC was evaluated for the first time using ETMT. The microstructure of the alloy after FAST, which is comprised of bcc Nbss and tetragonal αNb5Si3, was more homogenous compared with the cast alloy, with some interstitial contamination that occurred during powder production. The latter highlighted the importance of better control of processing conditions during milling of ingots. The FAST alloy had lower density than state of the art Ni-based superalloys and RCCAs and RHEAs, and its yield strength and specific yield strength was higher than those of the latter metallic UHTMs and comparable to those of RM (Nb)ICs and RM(Nb)ICs/RCCAs with B addition. The stress exponent n in compressive creep was in the range 1.7–2.6, similar to that of binary Nb–Si alloys and its creep rate at 1200 ◦C and 100 MPa was similar to that of the MASC alloy. Like the latter, the creep of the FAST alloy did not meet the creep goal.
Paper citation information
Journal of Materials Research and Technology 28 (2024) 3217–3224
Niobium-tungsten alloy Nb521 boasts excellent room-temperature and high-temperature mechanical properties, a high melting point, low density, high strength at high temperatures, and good machinability. Stardust Technology's spherical Nb521 alloy powder is produced using a radio frequency plasma spheroidization process. It features high purity, low oxygen content, high sphericity, a smooth surface, no satellites, uniform particle size distribution, excellent flow properties, and high bulk and tap densities. Niobium-tungsten alloy Nb521 is widely used in the manufacture of aerospace engines, weapon propulsion systems, liquid bipropellant engines for rockets and missiles, nuclear reactors, submersibles, gas turbines, automotive engines, diesel engines, high-temperature furnace heating zones, high-temperature molds, high-temperature fixtures, and high-temperature crucibles.
http://en.stardusttech.cn/products_det/109.html
For more information on niobium alloy powder products, please contact Vicky Zhang at +86-13318326185.
News
Stardust Technology (Guangdong) Co., Ltd.
101, Building 1, Liandong Youzhi Zone, Senshuji Road, Nansha Community, Danzao Town, Nanhai District, Foshan City,Guangdong Pro.,China
Tel：13318326187; 13318326185；
QR code