Platinum Metals Rev., 1978, 22, (1), 21
Order-Disorder Transformations in some Iron-Nickel-Platinum Alloys
Iron-platinum alloys have been extensively studied in recent years because of the shape memory effects that can be produced by suitable heat-treatment (1–3). Furthermore the formation of the thermoelastic martensites that underlie shape-memory effects seems to be linked to prior ordering in the parent austenite, and here iron-platinum alloys offer the possibility of varying this degree of order in a more systematic manner than most, if not all, other memory alloys investigated so far. One might have thought that iron-nickel alloys would be amenable to a similar treatment, since these two binary systems are analogous in so many ways; this turns out not to be the case, largely because ordering of bulk alloys at the analogous Fe3Ni composition seems to be absent. However, ordering in iron-rich iron-nickel alloys has been reported for powdered alloys, foils, and irradiated specimens (4–6), suggesting that its absence in bulk material may only be due to lack of diffusion. Since there is no particular reason to expect vastly different diffusion coefficients in iron-nickel and iron-platinum alloys, and since the ternary system iron-nickel-platinum has not been extensively investigated, it is plainly of interest to test a series of alloys along the quasi-binary section Fe3Pt−Fe3Ni with a view to obtaining data about (a) kinetics, (b) ordering, and (c) the characteristics of the martensite transformation in this area. Such a programme has been jointly organised, under the Science Research Council's Cooperative Award in Science and Engineering scheme, between the University of Surrey and the Johnson Matthey Group Research Laboratories, and some preliminary results of interest are reported here.
The observed variation of ordering temperature in ternary alloys with nickel-platinum ratio, together with a computer prediction of the values expected from a ternary phase diagram prediction programme developed by Inden (7), is shown in Figure 1.
It can be seen that the ordering temperature, as determined by both X-ray and differential thermal analysis techniques, drops from the platinum rich end of the section more or less as predicted, bearing in mind that the computer prediction uses an idealised set of parameters. However, as the midpoint composition is approached, ordering becomes increasingly more sluggish, leading to the absence of experimental data for ordering temperatures on the nickel rich side of the quasi-binary section (Figures 1 to 4).
This bears out the indications available in the literature concerning the effect of nickel, but it has not as yet proved possible to explain this change of behaviour. Figure 2 shows that at the platinum rich end of the system, an alloy containing 22.5 atomic per cent platinum and 2.5 atomic per cent nickel behaves very similarly to a 24 atomic per cent binary iron-platinum alloy, as evidenced by the lowering of the martensitic start (M.s) temperature with annealing time. Such a lowering of the M.s can be expected from ordering in the parent matrix, and should be accompanied by a sharp decrease in the hysteresis of the shear transformation, which is indeed borne out by the data in Figure 3.
However, with the addition of 5 atomic per cent nickel it is very clear that the onset of ordering has become very retarded (Figure 4) and by the time an alloy of composition 10 atomic per cent platinum and 15 atomic per cent nickel is reached, there is no indication of ordering from X-ray data (Figure 5). Although it is tempting to attribute this to lower diffusion rates, it can be seen from Figure 5 that marked changes in lattice parameter occur after annealing at lower temperatures in some of the more nickel rich alloys. For the present the only explanation of these anomalies—which are well outside the level of experimental error—is that there is some complex interplay between pre-ordering phenomena and the invar behaviour of this group of alloys (8–9). As far as the properties of disordered alloys are concerned, Figure 6 shows that there appear to be no great discontinuities across the section under investigation from the point of view of the lattice parameters of both austenite and martensite. Further work is in progress to relate the ordering behaviour with the invar effect before returning to the basic study of how these factors in turn affect the nature of the martensite transformation in this system.
- 1D. P. Dunne and C. M. Wayman Metall. Trans., 1973, 4, (1), 137
- 2T. Tadaki and K. Shimizu Scripta Metall., 1975, 9, (7), 771
- 3M. Foos, C. Frante, and M. Gantois in “ Shape Memory Effects in Alloys ”, ed. J. Perkins, (AIME Internat. Symp., Toronto, 1975 ). Publ. Plenum Press, New York, 1975, p. 407
- 4S. Kachi, Y. Bando and S. Higuchi Japan. J. Appl. Phys., 1962, 1, (6), 307
- 5G. Hausch and H. Warlimont Phys. Lett. A., 1971, 36, (5), 415
- 6J. Paulevé and D. Dautreppe J. Physique, 1963, 24, (1), 522
- 7G. Inden Max-Plank-Institute, Düsseldorf, private communication
- 8A. Kussman and K. Jessen Arch. Eisenhüttenw., 1958, 29, (9), 585
- 9K. Sumiyama, M. Shiga and Y. Nakamura J. Phys. Soc. Japan, 1976, 40, (4), 996