Investigating Variant Selection Criteria in Polycrystalline Shape Memory Alloys with In-Situ X-ray Topotomography

Predicting how shape memory alloys (SMAs) transform under mechanical loading is essential for designing reliable components in aerospace, biomedical, and actuator applications. A key factor in this is habit plane variant (HPV) selection—the determination of which martensitic variants are likely to form during transformation. While classical criteria such as maximum transformation work and resolved shear stress describe single-crystal behavior, they fail to capture the geometric constraints in polycrystals. Here, we propose a new HPV selection criterion that integrates both energetic driving forces and grain boundary compatibility. We find that variants most likely to form are those that maximize transformation work while minimizing grain boundary shear, promoting microstructural coherence across neighboring grains. This framework bridges the gap between theoretical predictions and experimentally observed stress-induced martensitic microstructures. To test this approach, we employ three-dimensional (3D) in-situ X-ray topotomography at the European Synchrotron Radiation Facility during tensile loading of CuAlNi and NiTi SMAs. This technique allows us to directly image, in 3D, the austenite parent grain and the evolving martensitic microstructures with sub-micrometer spatial resolution over millimeter-scale volumes. We observe direct visualization of variant selection across grain boundaries and at triple junctions. By connecting these observations with our proposed criterion, we establish an experimentally grounded framework for variant selection in polycrystalline SMAs, providing a pathway toward predictive microstructure design of SMA-based components.