Laser Annealing Induced Gradient Microstructure for Improved Fatigue Life and Elastocaloric Cooling Performance in Bending-Actuated NiTi Shape Memory Alloys

The elastocaloric effect of shape memory alloys (SMAs) offers a promising solid-state alternative to vapor-compression cooling. Among various actuation modes, low-force bending enables large reversible stress-induced phase transformation with reduced mechanical load and improved structural reliability. However, achieving uniform temperature distribution and long fatigue life remains challenging. This study introduces a laser annealing strategy to engineer gradient microstructures in NiTi SMAs by subjecting a commercial Ni₅₇Ti₄₃ flat wire (100x2x1 mm) to conventional furnace annealing followed by localized single-sided laser annealing across its thickness. Fine grains remained on the tensile side, while coarser grains formed on the laser-annealed compressive side. This gradient configuration improved transformation uniformity, as confirmed by IR thermal imaging and DIC strain mapping, and enhanced cyclic stability under continuous bending. During cyclic bending with maximum 7% compressive strain, the laser-annealed NiTi exhibited a maximum adiabatic temperature drop (ΔT_ad) of 10.2°C, approximately 18.4% higher than the as-received material, and a reduced energy dissipation (ΔW) of 118 N.mm. Under a 2.5% strain in a capstan-type continuous bending actuator, the fatigue life reached 5083 cycles, a 13% improvement over un-modified commercial nanocrystalline NiTi. DSC and XRD revealed a controlled shift in transformation temperature, increased latent heat and reduced internal stress concentration due to laser-induced recrystallization. The optimized gradient structure increased COP_material by raising ΔT_ad and lowering ΔW. This work demonstrates that post-heat treatment selective laser annealing can tailor microstructural gradients to achieve enhanced elastocaloric stability, higher cooling efficiency, and longer fatigue life under continuous bending actuation in SMAs.