Standard mechanical testing is inadequate for the mechanical characterisation of shape-memory alloys: Source of errors and a new corrective approach

Thanks to its unique behaviour characterised by a superelastic response, Nitinol has now become the material of preference in a number of critical applications, especially in the area of medical implants. However, the reversible phase transformation producing its exceptional comportment is also responsible for a number of phenomena that make its mechanical characterisation particularly complex, by hindering the assumptions at the very basis of common uniaxial tensile testing. This necessarily reduces the level of safety and design optimization of current applications, which rely on incorrect mechanical parameters. In this study, the spurious effects introduced by the unconventional material behaviour during uniaxial tensile testing are analysed by means of digital image correlation (DIC), identifying the onset of undesirable material inhomogeneities and bending moments that are dependent on the test setup and strongly limit the reliability of standard characterisation. Hence, a more accurate and systematic testing approach, exploiting the ability of DIC to analyse the local mechanical response at specific regions of the test specimen, is presented and discussed.