Characterizing the pseudoelastic behavior of Cu-Zn-Al shape memory alloys

Models of the pseudoelastic behavior of shape memory alloys (SMAs) have been developed by other researchers, but these do not incorporate the crystallography underlying the pseudoelastic effect, so they cannot predict SMA anisotropy. Single and polycrystal, models developed previously at the University of Virginia do incorporate the crystallography of the pseudoelastic effect and can therefore predict SMA anisotropy. The current research utilizes these models to investigate the impact of the crystallography of the martensitic transformation, sample grain size, and sample texture on pseudoelastic behavior. To assess the role of crystallographic features of the transformation, anisotropy and asymmetry indices are evaluated for a set of seven real and two synthetic SMAs, for both unconstrained and constrained single crystals. For the group Cu-Al-Ni, Cu-Zn-Al, Cu-Zn-Ga, CuZn, NiAl, the anisotropy decreases with increasing misorientation of the habit plane and shear direction from the ideal shear system orientation. This demonstrates the correlation between the crystallography of the transformation and the anisotropy of the pseudoelastic behavior for single crystals. These results are incorporated into an analysis of Cu-Zn-Al SMA polycrystals for which upper and lower bound stress-strain curves are predicted. Comparison with experimental stress-strain curves shows that samples with a small value of gs/t follow the upper bound transformation model, while samples with gs/t approaching one follow the lower bound model. These results strongly support the proposed explanation of constrained vs. unconstrained transformation for the observed effect of grain size on the stress-strain curves of Cu-based SMA. The stress-strain curve prediction capability is then extended to explore the impact of texture on pseudoelastic, behavior. Model stress-strain curves for Cu-Zn-Al and NiTi unconstrained and constrained polycrystals with and fiber texture of various strengths demonstrate that the effect of texture on the pseudoelastic behavior is dependent upon the specific texture components present, the state of grain constraint in the sample, and the alloy examined. Finally, a new model is developed to provide insight to the fundamental mechanisms of pseudoelastic behavior by predicting the type and volume fraction of martensite crystals formed when a SMA is loaded, which no previous model has done. Model results are in the form of calculated martensite pole figures, and these predictions correlate reasonably well with experimentally measured martensite pole figures from strained Cu-Zn-Al polycrystals. This, as well as a qualitative crystallographic analysis of the austenite to martensite transformation, provides confidence in the model predictions.