Volume 88, Issue 5

This paper provides a definition and classification of functionally graded materials (FGM). FGM manufacturing processes based on the modification of existing techniques like powder metallurgy are compared to non-conventional methods such as sedimentation. The ability of FGM’s to reduce thermal stresses is critically reviewed, and it is demonstrated that optimum performance can only be achieved if the thermomechanical property profile is adapted to the geometry and service conditions of the component. Determination of the thermomechanical property profile of FGM’s will require new techniques with high spatial resolution. The potential of FGM’s as high-temperature materials, hard materials, biomaterials and electronic materials is described and new applications in the field of functional materials are proposed.

General properties of magnesium alloys with one or two different rare earth metals are briefly summarized and discussed. The regular trends observed in the intermediate phase formation are used to predict the characteristics of the room temperature isothermal section of the Gd–Y–Mg system. The partial experimental section at 500 °C in the composition range ≈50 to 100 at.% Mg, has then been determined and found to be in fair agreement with the predicted one. The determination has been carried out by X-ray powder diffraction, optical and scanning electron microscopy and electron probe microanalysis. The following phases have been identified; the continuous solid solution cP2-CsCl type along the line Gd 1-x Y x Mg and, along the line Gd 1-x YxMg2, two Laves-type solid solution fields (cF24-MgCu 2 type, GdMg 2 based for about 0 ≤x<0.18 and hP12-MgZn 2 type, YMg 2 based for about 0.26 <x≤1). The extension into the ternary field, moreover, has been determined for Mg and the GdMg 3 , GdMg 5 and Y 5 Mg 24 based phases.

The letter is related to the recent Al–Co phase diagram by Gödecke and Ellner (Z. Metallkd. 87 (1996) 854) which contains a controversial point in comparison with our previously published diagram (J. Alloys Comp. 233 (1996) 279). We present arguments in favor of our diagram.

Phase equilibria were investigated for the Al-rich alloys Co 25-x Ni x Al 75 (0 ≤ x ≤ 25) and Co 22-y Ni y Al 78 (0 ≤ y ≤ 22). An addition of Ni to the Al-rich Co– Al alloys significantly decreases the binary peritectic, peritectoid and eutectoid temperatures. On the other hand, the temperature of the peritectic formation of NiAl 3 increases with increasing Co addition; the maximal solubility of Co in NiAl 3 amounts 10.5 at.% at 860 °C. Nickel substitutes Co in Co 2 Al 9 up to 4.7 at.%. The pseudoternary high-temperature phase (Co,Ni) 4 Al 13 (h) is stabilized by Ni addition up to 7 at.% Ni. The ternary phase Y 2 is formed by ternary peritectoid reaction from (Co,Ni) 4 Al 13 (h), Ni containing Co 2 Al 9 and the decagonal phase at 880 °C. It shows only a limited homogeneity range Co 17 Ni 8.2 Al 74.8 to Co 16.2 Ni 9 Al 74.8 .

Phase equilibria in the quaternary system Ti–Cu –Al –N are investigated using XRD, metallography and EDX, yielding the observation of 18 four phase spaces of the 850 °C isotherm. In the ternary boundary systems Ti– Cu–Al and Ti –Al –N, all previously reported phases are confirmed, but several tielines are newly determined. In the system Ti–N, the new ternary phase Ti 3 CuN dominates most phase fields. Based on thermochemical data of the binary boundary systems and the newly investigated or reinvestigated phase equilibria of the ternary boundary systems, a set of Gibbs energies for the solid phases occuring in the quaternary system is derived, which is used in combination with the quaternary experimental data to derive a large section of the isotherm for Ti –Cu –Al –N at 850 °C. The quaternary g-phase (Ti, Cu, Al) 6 N having the composition Ti 3 Cu 2 Al 1 N 0.8 is found to be the only stable quaternary phase of this system. Several η-phases isotypic to Ti 3 Cu 2 Al 1 N 0.8 are synthesized to explore the crystal chemistry effects of substitutions on different crystallographic sites.

The first stage of tempering of an iron–carbon–nitrogen martensite (containing 1.7 at.% C and 3.3 at.% N) was studied by atom probe analysis. Different stages of transformation were obtained by subjecting various specimens to different heat treatments. Analysis of the mass spectra revealed the presence of peaks not observed on decomposing binary FeC and FeN martensites: molecular ions occurred that contained both carbon and nitrogen. Upon tempering the FeCN martensitic specimens, interstitial-enriched regions develop. In most of these regions one type of interstitial is predominantly present but a small amount of the other type is present too. In combination with previously obtained X-ray diffraction results it was concluded that two types of transition precipitates are formed: nitrogen-rich regions also containing carbon and identified as α′′-like precipitates (α′′ is the transition precipitate formed in binary FeN martensite), and carbon-rich regions also containing nitrogen and identified as ε/η-like precipitates (ε/η is the transition precipitate formed in binary FeC martensite).

The influence of different heat treatments on a NiTi shape memory alloy with a nominal composition of 50.3 at.% Ti was studied with respect to mechanical and shape memory properties. The deformation behaviour both of martensitic condition B19′ at room temperature and high temperature phase B2 at 170 °C was investigated by tensile tests. The range of near-constant stress during deformation, the so-called martensite plateau, is more pronounced in the annealed samples. The yield strength of the high temperature phase strongly depends on the heat treatment condition. Annealing at 550 °C reduces it to about 50%. A two-way shape memory effect was induced by repeatedly heating and cooling under constrained conditions. Measurements of the transformation temperatures demonstrate that in specific cases the austenization during heating may occur in two steps.

In the precipitation-hardened superalloy IN 738 LC several deformation mechanisms, i.e. cutting of the γ′ phase or climb-assisted bypassing of precipitates by dislocations as well as the formation of dislocation networks at the γ – γ′ interfaces and particle coarsening, were observed depending on temperature and strain rate. For the description of the material behaviour a viscoplastic model based on the concept of internal stress is used. The internal stresses are attributed to particle strengthening and accumulation of dislocations. The used constitutive model is able to describe viscoplastic behaviour like the strain rate dependence and an increasing creep rate at small inelastic strain, i.e. the absence of a classical stationary creep range. Different hardening and softening rates of the model are active in different loading ranges. This can be explained by the operation ranges of different micromechanisms.

The applicability of the Tu –Turnbull and Zieba–Paw-lowski diffusion models of discontinuous dissolution have been considered. The simulation procedure applied to the Ni-1.4 at.%. In alloy has shown a much better applicability of the Zieba –Pawlowski model than the Tu–Turn-bull model for the prediction of the In concentration profiles left behind the receding reaction front. This comes from the fact that in the Tu –Turnbull model the same ratio between migration velocity and grain boundary diffusivity is assumed for both discontinuous precipitation and dissolution processes. This is only true when the difference between temperatures of discontinuous precipitation and subsequent discontinuous dissolution is relatively small. It has also been shown that the results of simulation become better with an increase of the residual supersaturation of the depleted α lamellae as well as with an increase of the characteristic solute concentration left at the places where the β lamellae have just been dissolved.

The interdiffusion coefficient (D̄) has been measured in the temperature range from 1442 to 1639 K as a function of the composition in the α (hexagonal close-packed, A3), α 2 (ordered hexagonal, D019), β (body-centred cubic, A2) and γ (ordered face-centred cubic, L10) phases of the Ti –Al system. The concentration profiles were obtained by electron probe microanalysis, and the chemical diffusivity was extracted from the concentration profiles by the Boltzmann-Matano procedure. In all the phases the interdiffusion coefficients increase with increasing Al contents. This increase is large in the disordered phase α, but much smaller in the ordered phases α 2 and γ.

The present work deals with an investigation of interdiffusion of cations in the binary oxide system MgO–“FeO” in the temperature range 1473 to 1603 K. Diffusion couple experiments have been carried out using polycrystalline samples of the two oxides and after the heat-treatment, the concentration profiles were measured using energy dispersive x-ray spectroscopy in a scanning electron microscope. From the results, the concentration and temperature dependencies of the interdiffusion coefficient have been evaluated as

Transmission electron microscopy (TEM) observation of in situ heated MgO–B 2 O 3 –SiO 2 slag allowed the description of its dynamic crystallization process. The crystallization processes are constituted by two steps: (i) phase separation and (ii) nucleation and growth of 2MgO · B 2 O 3 crystal precipitated prior to 3MgO · B 2 O 3 . The crystallization processes are sensitive to the electron beam. In electron irradiated regions, the phase separation occurred at 300 °C, however, the phase separation was observed in the electron non-irradiated region at about 750 °C.