Identifying the Intermetallics for Investigation

Although many physical parameters affect the mechanical properties of a prospective material, if many materials are to be surveyed it is useful to decide on a few meritorious properties, for which data are generally available, in order to choose materials for initial screening. In this instance the candidates are numerous; a search for intermetallic compounds that melt above 1500°C identified nearly 300 such binary compounds (1).

Two basic physical properties were identified to simplify the screening, these being the melting temperature, T_m, and the specific gravity, ϱ. Because these properties are insensitive to processing variables and the resultant microstructure, and are only slightly altered by minor variation in alloying elements, they are helpful indicators of several vital properties. The melting temperture is by far the most useful structure-insensitive property. First, it specifies the temperature interval over which materials are solid. Second, its magnitude increases with the stiffness of a material, since the elastic moduli correlate strongly with melting temperature (2). In addition, all models of strengthening give values of flow stress that are proportional to elastic constants, which in turn increase with the melting temperature. Fourth, expansion coefficients (which for convenience should be small) vary inversely with melting temperature. Finally the creep rates define a maximum operating temperature that increases with melting temperature. Approximate limits on operating temperatures for single-phase materials are estimated to lie between T_m/2 and 2T_m/3, with T_m/2 being more typical (3).

For aerospace application, and in rotating parts, high specific strength, that is strength per unit density, and specific stiffness are important. Thus the specific gravity enters consideration, and materials with high melting point and low specific gravity are needed.

The choices can be clearly identified by displaying the data on graphs which plot specific gravity versus melting point, using different charts for different crystal structures. Figure 1 is particularly relevant, since it includes all four of the compounds of primary interest here (1). It surveys two of the simplest ordered structures, the B2 CsCl-type cubic lattice, with 2 atoms per unit cell, and the Ll₀ AuCu-type tetragonal lattice, which has 4 atoms per unit cell.

The desired combination of properties occurs in the lower right region of such a diagram. Since no compounds he there, one works from the lower right towards the upper left for the best prospects. Where specific gravity is over-poweringly important, AlTi, AlNi, and AlCo would be early choices. Of these AlTi and AlNi are already the subjects of intensive activity; AlCo is not, however, because it is extremely brittle. RuSc and AlRu are obvious subsequent choices because of their high melting points.

Preliminary Screening Tests

Simple mechanical tests were selected to characterise strength versus temperature, and to assess room-temperature toughness. Extensive microhardness measurements were carried out, the results giving an indication of strength (4). An absurdly simple “chisel toughness” test was devised, which consisted of applying to samples a chisel and two hammers of different weights in sequence, to define a 4-level toughness scale that is qualitative, but reproducible (4). The highest level, chisel toughness 3 (unbreakable by this test), was attained by the four compounds listed earlier, but by no other single-phase binary intermetallic.

In all, 20 crystal structure types, 90 binary compositions, and 130 ternary or higher order alloys have been studied by such tests (5). The elastic constants were determined for most of them by more conventional techniques.

The brittleness of most high-temperature intermetallics is such that at room temperature a compression test will give no measurable plastic strain prior to fracture. The results presented in Figure 2 are strikingly different, showing limited but definite plastic strains for the two Ll₀-based alloys and large plastic strains for the two B2-based alloys. These special properties drew attention to these compounds and led to further studies of them and their alloys (6–9).

Tough Alloys

Some characteristic properties of the four tough compounds are listed in the Table (5). Most noteworthy are the high melting point values, which place these compounds as a group in the top 17 per cent of all the binary intermetallics that have melting points above 1500°C. Except for RuSc, the moduli are also high, in the top 33 per cent of high-temperature intermetallics.

IrNb

Although IrNb is expensive, it is less so than pure iridium, which at present is used in the radioisotopic thermoelectric power sources employed on satellites in orbit. That is one potential application; another, with a much lower probability for use, is for jet engine blades or disks. The latter is only likely to occur if a significant part of the iridium can be replaced without seriously degrading the properties of IrNb. To date, the replacement attempts—12 alloys with cobalt, niobium and nickel—have not been successful (6). Also, oxidation problems exist, so that coatings would be needed (10).

Much of the toughness is suspected to arise from the fine-scale internal structure that is common—twins in some cases (see Figure 3), and alternating IrNb Ll₀ and Ir₁₁Nb₉(oP12) orthorhombic lamellae in others (Figure 4).

RuTa

If RuTa can be developed further, and if lighter alloys such as those based on AIRu do not develop adequately, then RuTa-based alloys might be considered as jet engine materials. They have, however, major problems, these being: oxidation (10), which might be counteracted by coating, and high density, which can be disregarded only if these alloys allow engine temperatures to be significantly increased. Some twenty-two RuTa-based alloys have been tested, and hardness, toughness, and elastic properties reported (7, 9). Of these, nine were in the highest toughness categories, and an additional seven had toughness equal to that of Ti₃Al and TiAl (9), two lower melting point alloys that are well on their way to finding applications in engines or in space.

Replacements of a portion of the ruthenium by cobalt or iron could be made while still retaining most of the room temperature toughness; sample compositions are Co₂₀Ru₃₀Ta₅₀ and Fe₁₅Ru₃₅Ta₅₀. Elastic moduli are decreased by changing the ruthenium content away from 50 atomic per cent, as shown in Figure 5. Single-phase alloys follow the lines drawn while two-phase alloys have considerable scatter.

RuTa, like the other Ll₀ crystal type IrNb, has a fine structure of twins (Figure 6) or alternating lamellar phases (in this case Ll₀ and B2) (7). As Figure 7 shows, the layer structure appears clearly on fracture surfaces. Plasticity and toughness are thought to be strongly affected.

AIRu (and RuSc)

For aerospace applications the B2 compounds appear to be more promising than those with the Ll₀ structure because of their greater ductility and lower specific gravities; and AIRu suffers the least from oxidation. The data imply that uncoated Al₄₇Ru₅₃ and Al₄₈Ru₅₁Y could be used at temperatures up to 1250°C (10). Scandium makes pure RuSc unusually expensive, but the fact that RuSc has good mechanical properties has been used to infer, rightly, that additions of scandium, replacing some of the aluminium in AIRu might result in improved properties (8, 9).

Many AlRu-based materials have been tested. Among several dozen compositions, twenty-one were designed to test solid solution hardening by cobalt, iron, and titanium, twenty-two were used to explore off-stoichiometry effects and the effects of boron and scandium on toughness and strength, and eight were concerned with the effects of adding a variety of elements (cerium, chromium, rhenium, silicon and yttrium) which were intended to lower oxidation rates.

The ruthenium concentration had a powerful effect on the elastic properties of RuTa in the composition range of 43–60 atomic per cent; here Young's modulus was raised by about 1 per cent for each per cent increase in ruthenium content. Among the other alloying additions only scandium exerted a significant effect, which was of lowering the moduli by about 1 per cent for each percentage addition.

Stoichiometry and minor additions of boron have profound influences on the plasticity of AlRu-based materials, as Figure 8 shows. Compounds that are sub-stoichiometric in ruthenium are brittle at grain boundaries, but the plasticity of AIRu is clear, and both work hardening and strain-to-fracture increase with the fraction of ruthenium. Also apparent is the beneficial effect of adding 0.5 per cent of boron, which doubled the plastic strain and maximum stress. The boron is also effective at that level for sub-stoichiometric and stoichiometric alloys. Figure 9 shows this behaviour, and also that the particular concentration of 0.5 per cent boron was a happy choice, since both lesser and greater levels were ineffective.

The scanning electron micrographs of fracture surfaces presented in Figure 10 show the striking change in surface character that results from increasing the ruthenium content from 47 to 53 per cent and from adding 0.5 per cent boron. Al₅₃Ru₄₇ is intergranularly brittle; Al₄₇Ru₅₃ + 0.5 per cent boron is tough, with a fibrous fracture surface. The intergranular brittleness and its cure by boron addition is reminiscent of the situation in Ni₃Al (11), the base material for nickel-based superalloys, which are the present materials of choice for jet engine blades and disks.

Scandium was thought to be a possible solid solution substitution for aluminium since RuSc and AIRu have in common the B2(cP2)CsCl structure (12), and since their lattice constants (3.1985 ± 0.0006 Å and 2.9916 ± 0.0007 Å, respectively) differ by only 6.9 percent. Nevertheless, even minor scandium additions led to the appearance of a second phase (8), but one with helpful effects on the high-temperature strengths as inferred from microhardness measurements. However, 10 per cent scandium or more greatly decreased the toughness (8), but the use of boron to improve ductility and toughness was again successful for compositions with lower scandium content. Figure 11 compares the hardness of stoichiometric AIRu with that of two alloys, one containing 5 per cent scandium and the other 4 per cent scandium and 0.5 per cent boron. Both have improved hardness to 1150°C, and the boroncontaining alloy has outstanding toughness. Another alloy with 2 per cent scandium and 1.5 per cent boron was nearly as tough.

Ternary Alloys

Recently both toughness and tensile ductility have been found by Waterstrat, Bendersky and Kuentzler in a ternary B2(cP2) structure, Zr(Pd_0.7Ru_0.3), which contains two platinum group metals (13). Although cracks formed at about 1 per cent tensile strain, macroscopic fracture occurred at greater than 6 per cent strain. Slip with <100> slip vectors was accompanied by the additional deformation mode of {114} twinning, which was suggested to be the mechanism that allowed plasticity. The melting point is bracketed in the range as 1800 ± 200°C.

Overview and Potential Uses

As noted in the second paragraph of this article, cost is one major obstacle in the use of ruthenium-based materials in jet engine components. In part, costs might be reduced in the future by partially replacing ruthenium with lighter, less expensive elements. Prices may decrease if more massive demand leads to reduced costs of production or to discoveries of new ore. Improvements in advanced jet engines are driven by military demands, which are much less governed by cost than are commercial customers. Therefore, military engines are the first candidates for engine applications. A final critical obstacle to use is that although the preliminary properties are encouraging, many rigorous tests still need to be passed before it is clear whether these materials are technically capable of being used in engines. As R. Sprague of the General Electric Company has noted, “The first thing you ever hear about a new material is usually the best thing you ever hear.”

The possible use of IrNb alloys in place of iridium in radioisotopic thermoelectric generators is an even more specialised application, and it remains speculative.

Zr(Pd_0.7Ru_0.3) is being evaluated as a dental alloy, where cost is a minor consideration.

In a personal communication Dr. M. B. Cortie has noted that ruthenium commonly aids corrosion resistance, and that AlRu-based alloys may therefore be useful in applications where corrosion resistance at unusually high temperatures is vital.