Non-generic concentrations for shape-memory alloys; the case of CuZnAl

As is well known, the remarkable properties of many active crystalline materials, such as shape-memory alloys, originate from a symmetry-breaking martensitic phase transformation. A recent analysis (see [1]) has shown that if a special condition on the strain parameters is satisfied, the twinning ability of a crystal undergoing a cubic-to-monoclinic transformation is considerably increased. As this feature may improve the memory behavior of cubic-monoclinic alloys, we suggest that these predictions be tested experimentally. In this paper we give a procedure to determine explicitly which ‘non-generic’ concentrations allow suitable alloys to produce the extra twins; we also determine which concentrations give a volumepreserving transformation (this condition guarantees self-accommodation of the martensite). As an explicit example, we consider the ternary CuZnAl system, for which, based on the available experimental data, we establish the non-generic concentrations. Our work indicates that there is a scarcity of the necessary data on the mapping of lattice parameters versus concentrations in alloys with a potential for shape memory. We propose a specific experimental program aiming at the systematic gathering of this information: building such a database may prove of considerable help in the search and synthesis of new materials with improved performance.