Martensitic Transition

The body-centered-cuhic (bcc) metals and alloys are normally classified as undesirable for low temperature construction. This class includes Fe, the martensitic steels (low carbon and the 400-series stainless steels). Mo, and Nb. If not brittle at room temperature, these materials exhibit a ductile-to-brittle transition at low temperatures. Cold working of some steels, in particular, can induce the austenite-to-martensite transition. 

On the other hand, the formation of the high pressure phase is preceded by the passage of the first plastic wave. Its shock front is a surface on which point, linear and two-dimensional defects, which become crystallization centers at super-critical pressures, are produced in abundance. Apparently, the phase transitions in shock waves are always similar in type to martensite transitions. The rapid transition of one type of lattice into another is facilitated by nondilTusion martensite rearrangements they are based on the cooperative motion of many atoms to small distances. ... 

In contrast to this the austenite - martensite transition temperature depends on the concentration of vacancies. The configurations A and E show a transition temperature of 10 K, while in C and D the bcc structure already occurs at 100 K. 

The transition temperatures which we find for the austenite - martensite transition in simulations without vacancies, are much to low. The introduction of vacancies lead to much more physical results, while the martensite - austenite transition is not affected by vacancies. From this we conclude that vacancies lower the energy barrier which the system has to overcome during the transition. The reason for this might be a weakening of long-range elastic couplings in the lattice. 

We like to conclude this introduction observing, after Lifshitz, that a free energy exeess is involved in each ETT. TTius, it might happen that, in order to avoid the energy cost involved in opening, say, a neck in the FS, the system might prefer to change its phased. This could be the case of the martensitic transition observed in CuPd. 

D.J. Gunton and G.A. Saunders, The Elastic Behaviour of In-Tl alloys in the vicinity of the martensitic transition, Solid State Commun. 14 865 (1974). 

Phase transitions in solids are also fruitfully classified on the basis of the mechanism. The important kinds of transitions normally encountered are (i) nucleation-and-growth transitions (ii) order-disorder transitions and (iii) martensitic transitions. 

Hyde, S. T. Andersson, S. 1986 The martensite transition and differential geometry. Z. Kristallogr. 174, 225-236. 

Fig. 4. 2-dimensional pictorial representation of atomic movements in a martensitic transition. 

The existence of acoustic emission (AE) during the martensitic transition in various systems is well known [7,8,9]. However, in Nitinol the AE is missing as shown in the private communication I had with Dr W.F. Hartman of John Hopkins University as shown in the excerpt below ... 

This unique property is actually related with the unique property of 1.1. Because the reason for the AE in martensitic transition is that the transition takes place in steps (as illustrated in Fig. 4) i.e., the whole two-dimensional plane of atoms shifts altogether at a time. This abrupt shift results in audible click sound. On the other hand, Nitinol does not undergo such shift but rather undergo a continuous shearing (as illustrated in Fig. 3) so that there is no audible sound. 

As is explained above, single crystals of size within a fraction of a mm (millimeter) are not available in Nitinol alloy (under microscope the grain size was determined to be in 100 micron range). On the other hand, other conventional martensitic transition systems, such as In-Tl, Au-Cd, Al-Cu all have good size of crystals — in the range of 1 4 cm. 

One of the well-known properties in the martensitic transition is that it undergoes macroscopic change in shape. This can be visualized by the illustration of Fig. 4. In Nitinol, no macroscopic shape change is observed. This can also be visualized by Fig. 3, in which the atomic shears take place in zigzag fashion. Since the atomic shears are all within interatomic distances macroscopically no shape change is observed. 

It is well established that all the known martensitic transitions, before Nitinol, occurs through crystallographic transformation [11], As described above [10], single crystal X-ray diffraction work shows a crystallographic distortion (rather than transformation) and therefore is a second-order transformation [12], This conclusion has the support of heat capacity investigation [13] that gave a latent heat of transition AH >... 

But, aside from these unique properties, Nitinol has a number of commonalities with other known martensitic transition systems (1) it is an athermal transformation, (2) it is diffusionless, (3) it involves displacive or shear-like movement of atoms, (4) the activation energy for the growth of martensite (continuous atomic shear in Nitinol) is effectively zero, i.e., the propagation rate of transformation (transition in Nitinol) is fast and independent of temperature. 

It was based on these uniqueness and commonalities, my colleague and I submitted a paper entitled Crystal Structure and A Unique martensitic Transition of TiNi to a Journal concerned with metals and alloys for publication in 1965. But, the paper was rejected outright by two anonymous reviewers who could not accept our observation that the Nitinol transition was unique. Obviously the reviews contend that by accepting Nitinol transition being unique, may make all other martensitic transformations garden variety. This may upset the theory of martensitic transition formulated thus far. We then, submitted the paper to the Journal of Applied Physics and was accepted for publication and eventually appeared in print [10]. A few months after the appearance of this article, the editor of the very journal that rejected my paper, asked me to review two papers on Nitinol for the journal. Suddenly, I was an undisputed expert in Nitinol Up to this point I had not really start to apply covalent-bond concept but devoting more time in collecting experimental data [14,15], which may be important in support or non-support of covalent-bond concept. 

In 1967, on April 3 and 4, under the sponsorship of ONR (Office of Naval Research) I organized the first International Conference on Nitinol called Symposium on TiNi and Associated Compounds . The conference was held at Naval Ordnance Laboratory, the birthplace of Nitinol. As the chairman of the conference I assisted in selecting the papers from this conference that were later published in block form in the Journal of Applied Physics [16]. Despite these efforts the Nitinol transition remained elusive for sometime. In fact, after more than 30 years since the discovery of memory effect and with more than 139 papers have appeared in various journals on this subject, the investigators still do not agree with one another. At the same time more than 4,000 patents worldwide have been filed on the use of the memory effect or superelasticity in Nitinol. Out of all this, the actual application of Nitinol remains only a handful. In sharp contrast other conventional alloys with martensitic transition has no controversy and in fact they are so well understood that a Crystallographic theory of martensitic transformation was formulated [27],... 

Fig. 12 Electrical resistivity of TiNi(51 at. %Ni) at around the martensitic transition temperature effect of thermal cycling in (a) no incomplete thermal cycling, (b) after six incomplete thermal cycling, (c) after several hundred incomplete thermal cycling. Complete and incomplete thermal cycling are defined in the text.

Fig. 20. Density of states of TiNi Ts and Tf are the upper and lower temperature limits in the TfNi martensitic transition, (a) Density of states below the transition. Note that the llower d-band edge is at 0.17 eV. The s-band contains 1.5 electrons, whereas the d-band Contains 5.5 electrons, (b) Density of states above the transition. The lower d-band edge is at 0 eV. The s-band and d-band contain 0.1 and 6.9 electrons respectively.

By recognizing the existence of covalent bonded electrons in metals and alloys I have shown how it is possible to explain and understand simultaneously a variety of physical phenomena observed, ranging from Phase Diagrams to Superconductivity to Mechanical Properties to Martensitic Transition in Nitinol. In this aspect I personally derived a tremendous gratification. It is my fervent hope that the material presented here will touch the hearts of the reader in some small way and will make them pause to think the importance of electrons, particularly the covalent bonded electrons, in the world of metals and alloys. 

The structural (martensitic) transition in Ni2MnGa-based Heusler alloys was described as driven by a band Jahn-Teller effect [ff, 12], In the framework of this model, the results of theoretical [12] and experimental [If] studies of the redistribution of magnetization upon martensitic transformation in Ni2MnGa alloys have been discussed. 

Figure 6. Magnetization jump at the martensitic transition in magnetic fields of 3 and 5 T. For 0 < x < 0.16 compositions M(T) dependencies were measured upon heating. For the x = 0.19 composition, a temperature hysteresis loop of the magnetization observed at martensitic transition is shown.

Figure 7. The magnetization jump at the martensitic transition in various magnetic fields as a function of Ni concentration in NT+sMni-sGa (0 < x < 0.19) alloys.

Figure 9. Compositional dependence of the latent heat of the martensitic transition...

---