Mechanical Characterization of Additively Manufactured Functionally Graded NiTi Shape Memory Alloys

NiTi Shape Memory Alloys (SMAs) are highly valued for their unique Superelastic (SE) properties, Shape Memory Effect (SME) and recoverable deformation. However their application is often limited by functional fatigue and the accumulation of residual strain in cyclic loading, compromising their cyclic stability. 

The performance of NiTi SMAs relies heavily on the microstructure and composition, both of which are strongly affected by the manufacturing process. 

Additive Manufacturing (AM) and specifically Laser Powder Bed Fusion (LPBF), offers an unprecedented ability to tune the microstructure, and thus the properties and performance of NiTi SMAs, by adjusting the process parameters.

It also offers the possibility to fabricate NiTi components with complex geometries and implement techniques such as Functional Grading (FG), where two different microstructures co-exist in the same sample to give rise to novel functional behaviours. 

This thesis aims to investigate how processing route and microstructure govern the compressive superelastic response and cyclic degradation of NiTi, with emphasis on LPBF manufacturing. 

Homogeneous samples fabricated with LPBF (using various sets of process parameters), casting and rolling, were mechanically characterized. Furthermore, another FG sample was fabricated with LPBF, where the core had strong texture preferable for superelasticity, and non-textured outer sections. 

All samples were subjected to cyclic compression tests at austenitic state (80 °C) to ensure superelasticity, with in-situ Digital Image Correlation (DIC) setup to map strain evolution. LPBF samples produced measurable differences in texture/defects that resulted in distinct stress-strain loops and strain localization. A4 with strong texture showed signs of early slip activation, while A6 with weaker texture and porosity promoted stress localization and faster cyclic degradation. 

All homogeneous samples accumulated over -2% residual strain. The rolled sample exhibited high stresses and eventually buckling, with the deformation being accommodated by mechanical twinning. Lastly, the cast sample exhibited early functional fatigue in the first 3 cycles, and localized deformation in the top section as shown in optical micrographs.  

In contrast the FG sample exhibited superior cyclic stability accumulating only 0.89% of residual strain after 10 compression cycles. The DIC stain maps showed that most of the strain was carried out by the core, while the outer sections remained comparatively elastic and shared load. Optical micrographs also revealed that irreversible damage concentrated at the interface between the textured core and non-textured outer sections. FG sample was also tested to 6 compression cycles at 6 different temperatures (60-110 °C), where at lower temperatures it behaves more homogeneously while at higher temperatures the core carries most of the strain. 

The results of this work demonstrate that LPBF process parameters can be used to tailor compressive superelasticity and cyclic stability and that functional grading of crystallographic texture is a highly effective strategy to mitigate cyclic degradation of NiTi, enabling the design of durable and high performance NiTi components.