Machine learning guided alloy design of high-temperature NiTiHf shape memory alloys

Abstract With the increasing use of CubeSats in space exploration, demand for reliable high-temperature shape memory alloys (HTSMA) continues to grow. A wide range HTSMAs has been investigated over past decade but finding suitable by means trial-and-error experiments is cumbersome and time-consuming. The present work uses a data-driven approach identify NiTiHf actuator applications space. Seven machine learning (ML) models were evaluated, best fit model was selected new alloy compositions with targeted transformation temperature (Ms), thermal hysteresis, output. Of studied models, K-nearest neighbouring ML offers more accurate prediction developing balanced functional properties aids our existing understanding on compositional dependence temperature, hysteresis For instance, sensitive Ni variation Hf content. maximum Ms reduction rate 6.12 °C per 0.01 at.% attained at 30 Hf, content between 50 51 at.%. Graphical abstract