Data-driven study of composition-dependent phase compatibility in NiTi shape memory alloys

The martensitic transformation in NiTi-based Shape Memory Alloys (SMAs)\nprovides a basis for shape memory effect and superelasticity, thereby enabling\napplications requiring solid-state actuation and large recoverable shape\nchanges upon mechanical load cycling. In order to tailor the transformation to\na particular application, the compositional dependence of properties in\nNiTi-based SMAs, such as martensitic transformation temperatures and\nhysteresis, has been exploited. However, the compositional design space is\nlarge and complex, and experimental studies are expensive. In this work, we\ndevelop an interpretable piecewise linear regression model that predicts the\n$\\lambda_2$ parameter, a measure of compatibility between austenite and\nmartensite phases, and an (indirect) factor that is well-correlated with\nmartensitic transformation hysteresis, based on the chemical features derived\nfrom the alloy composition. The model is capable of predicting, for the first\ntime, the type of martensitic transformation for a given alloy chemistry. The\nproposed model is validated by experimental data from the literature as well as\nin-house measurements. The results show that the model can effectively\ndistinguish between $B19$ and $B19^{\\prime}$ regions for any given composition\nin NiTi-based SMAs and accurately estimate the $\\lambda_2$ parameter. Our\nanalysis also reveals that the weighted average of the quotient of the first\nionization energy and the Voronoi coordination number is a key compositional\ncharacteristic that correlates with the $\\lambda_2$ parameter and thermodynamic\nresponses, including the transformation hysteresis, martensite start\ntemperature, and critical temperature. The work herein demonstrates the\npotential of data-driven methodologies for understanding and designing\nNiTi-based SMAs with desired transformation characteristics.\n