Functional thermomechanical responses of NiTi due to the martensitic transformations are frequently utilized in engineering applications and stress–strain temperature constitutive behavior is simulated by SMA models. In both cases, it is frequently implicitly assumed that the involved strains are completely recoverable and the responses are stable upon cycling. In reality, however, this is not the case. On the contrary, the cyclic thermomechanical behaviors of NiTi are inherently unstable, the concern is mainly how much. The drift of the stress–strain-temperature response during cycling is termed functional fatigue. It is considered to be a problem because of two reasons: since it adversely affects the performance of NiTi elements in engineering applications(superelastic members, vibration damping members or actuators) and since it is presumably responsible for fatigue degradation and preliminary failure of NiTi. Impact of the functional fatigue on practical applications can be partially eliminated through stabilizing the cyclic responses by training but the limited fatigue life remains to be a serious problem for the SMA technology. The link between the functional fatigue and structural fatigue can be possibly understood in terms of dissipation energy based criteria for structural fatigue (the more energy is dissipated during the cyclic stress induced martensitic transformation, the shorter is the fatigue lifetime). Given the growing evidence on the role of environmental effects in superelastic fatigue, however, this cannot be safely considered as a general controlling mechanism of the Nitinol fatigue, particularly to the superelastic deformation in fluids..
A bicrystal model for “microstructure evolution” during tensile superelastic cycling of polycrystalline NiTi wire. Tensile load axis is oriented differently with respect to the austenite lattice in grains A and B of the NiTi bicrystal (1). Stress induced martensitic transformation in grain B proceeds at room temperature at lower stress and is accompanied by more dislocation slip due to the anisotropy of the hybrid slip/transformation deformation mechanism. Tensile deformation of grains A, B is constrained by the vertical grain boundary. When the model bicrystal undergoes stress induced martensitic transformation (2), internal stress is lower in grain B than in the grain A due to the anisotropy of the hybrid slip/transformation deformation mechanism. After unloading (3), dislocation defects (grain B), incremental plastic deformations (grain B), residual stress and residual martensite variants (tensile in grain A and compressive in grain B) appear as suggested in 3b. Upon further cycling (4), the microstructure evolves and internal stress continuously redistributes, which brings about the instability of cyclic superelastic stress–strain response of the bicrystal, particularly to the accumulation of macroscopic nonrecovered strain, defects, residual martensite and drift of the forward and reverse transformation stresses (compare 2 and 4).