Materials Science and Technology

A metallographic investigation of splat-cooled M1 high-speed steel is described. Apart from general microstructural refinement, the splat-cooling process retains in solid solution much greater amounts of the alloying elements than conventional quenching processes. This stabilizes the austenite to such an extent that the M _s temperature is depressed close to or below ambient temperature and no martensite is formed. The morphology of the matrix phases, δ-ferrite and austenite, is shown to depend critically upon the specimen thickness and hence local cooling conditions. Owing to the high solute supersaturation, subsequent tempering of the splat-cooled steel leads to precipitation of the vanadium-rich carbide MC on 100 planes in the δ-ferrite, in sharp contrast to the molybdenum-rich M₂C carbide which is normally responsible for secondary hardening in conventionally processed high-speed steels. In the splat-cooled steel, M₂C appears only at the solute-rich interdendritic boundaries. Sintering studies were carried out to consolidate splat-cooled material into a more useful form. The presence of high concentrations of carbon in solution in the as-splatted material depresses the solidus temperature so much that partial melting occurs upon sintering at 1200°C, while the excess dissolved carbon permits the precipitation of M₂C at 1100°C. The practical application of the knowledge gained from this study to possible future developments in high-speed steel technology is discussed.

Magnesium specimens containing aluminium, zinc or silver have been tested in tension after air cooling or water quenching from temperatures in the solid solution range. All materials exhibited an increment in flow stress (compared with air-cooled specimens) after water quenching that tended to increase with quenching temperature and alloy content. Variations in the flow stress on subsequent ageing at temperatures up to 280°C are consistent with hardening by the precipitation of further loops of dislocation and subsequent softening as vacancies are lost to sinks.

Whatever the heat-treated condition or composition of the specimens, the flow stress increase over the air-cooled value can be interpreted in terms of the concentration and size of dislocation loops, and all alloys display a linear dependence on the square root of the product of loop density and diameter until the mean distance between loops decreases to less than six loop diameters. Beyond this point there is a more complex dependence upon the same parameters.

Creep deformation and rupture have been studied in 2¼Cr-IMo steel over the stress range 60-210 MN m⁻² at 565°C. Creep damage accumulates by the initiation and growth of extensive cavitation at the prior austenite grain boundaries. Cavity formation predominates during the initial transient and individual cavities appear to nucleate on grain boundary carbides. Quantitative analysis of the cavitation kinetics in relation to the creep deformation processes suggests that cavity growth is directly related to deformation occurring at the grain boundaries. It is inferred that the vacancy diffusion mechanism is inhibited and growth is limited by the local creep process occurring at the grain boundaries. At low stresses the latter is controlled by intra-granular recovery creep such that the overall rate of cavitation is proportional to the macroscopic strain rate. However, results obtained under biaxial shear conditions imply that the rate of cavitation is additionally dependent upon the normal boundary stress or maximum principal stress. At very high stresses evidence indicates a transition in rupture process to one of ductile shear. The implications of these results are discussed with regard to high-temperature plant integrity.

With Nimonic 105, the dependence of the secondary creep rate ε_s on stress σ and temperature T can be described as

ε_s = Aσⁿ exp - Q _c/RT

with n ≃ 7 and Q _c ≃ 515 kJ mol⁻¹. Using a technique which involves making consecutive small stress reductions during creep, definitive values of the friction stress σ₀ can be determined such that these large values of n and Q _c can be rationalized as

ε_s = A*(σ-σ₀)⁴ exp - Q _c */RT

where Q _c *, derived from the temperature dependence of the creep rate at constant (σ-σ₀), is equal to the activation energy for creep of single phase nickel–chromium alloys (∼ 310 kJ mol⁻¹). This suggests that processes occurring in the matrix are rate-controlling during creep of nickel-base superalloys. Differences in the stress and temperature-dependence of creep resulting from variations in creep testing procedure are shown to be attributable to micro-structural instability. Furthermore, measurements of σ₀ during conventional creep tests and during tests in which the specimens were subjected to creep/heat-treatment cycles demonstrate that, under the conditions studied, tertiary creep commences as a result of precipitate coarsening rather than as a consequence of grain boundary cavity and crack development. Improved creep rupture lives can then be achieved by periodically reheat-treating the material to restore the original particle distribution.

Auger electron spectroscopy (AES) has been used to detect and determine the depth distribution of minute quantities of boron in the surface layer of Fe-10%Cr steel oxidized at 600°C in air contaminated with vapour phase boron. Such minute quantities of boron, affording remarkable corrosion protection, are shown to be concentrated in the surface layer 10–20 nm thick and associated with a large enhancement of chromium. Mechanisms for the boron inhibition effect are discussed in the light of these new AES results. The applications in which the protection may be utilized in a service environment can now be more closely delineated.

Analysis of engineering steels by Auger electron spectroscopy has identified phosphorus, and to a lesser extent tin, as the impurities which commonly cause temper brittleness. Following a thorough thermochemical assessment of the feasibility of ‘gettering’ these residuals by forming stable microcompounds, the rare earth elements lanthanum and (to a lesser extent) cerium were selected as having the best promise because of their higher affinity for P and Sn than for the carbon of the steel.

Measurements on the ductile-brittle transition temperatures were carried out on two archetypal steels, 2¼Cr-1Mo and 3½Ni-Cr-Mo-V, in both high purity and commercial purity conditions. In the 2¼Cr-1Mo steel, P was shown to be very embrittling but Sn had little effect. The embrittling effect of as much as 0.02%P could be relieved by an addition of about 0.16%La. Concomitantly, there occurred the detrimental effect of a reduction of upper shelf energy in the ductile region. This reduction is thought to have arisen from the small laboratory-scale experiments and is not expected for industrial practice. The behaviour of the 3½Ni-Cr-Mo-V was similar to the 2¼Cr-1Mo except that both P and Sn were very embrittling, and the optimum La remedial additions were slightly lower. In both types of steel the phosphorus and tin temper brittleness can be fully removed by an optimum lanthanum addition, given by La = 8.7S + 2.3Sn + 4.5P, where the element symbols express concentration in wt-%. Tests with mischmetal showed an equivalently effective remedial action.

The oxidation resistance of mild steel is shown to be considerably enhanced by additions of nitrogen which result in the formation of an adherent fine-grain-size oxide scale. A mechanism is proposed by which improved protection results from enhanced nucleation of oxide at incoherent nitride particles in the surface metal layer. A suitable nitrided surface is forme by constant activity ageing in an NH₃ : H₂ gas mixture to form austenite at nitriding temperature. The resulting austenite + martensite micro-structure tempers rapidly during oxidation at 420°C forming incoherent Fe₄N precipitates which act as oxide nucleation sites. Similar results have been obtained from studies of iron alloys containing small additions of titanium. Formation of a surface layer containing a fine dispersion of titanium nitride particles in a ferrite matrix is shown to enhance oxidation resistance.

The proposed mechanism is similar to that of the development of improved oxidation resistance in superalloys by subsurface dispersion of incoherent stable oxide particles.

Current technological interest in the intermetallic compound ε-FeTi stems from its potential use for hydrogen storage. This note concerns the effect of small additions of cerium on the hydrogen-storage capacity of the compound.