Intermediate porosity (25%-40%) NiTi is processed by Spark Plasma Sintering (SPS) method. In order to increase the porosity, the SPS chamber setup is modified. A pair of spacers is added to the chamber punch setup so that the pressure applied on the powder is minimized. As a result, the porosity is increased. TiH<sub>2</sub> powder is added as a pore-forming reagent to the element powder mixture for sintering. The decomposition of TiH<sub>2</sub> increases the porosity efficiently. Two kinds of heat treatments are applied to convert the porous NiTi to superelastic grade. One is homogenization followed by aging treatment, the other is aging treatment only. It is found that homogenization heat treatment reduces the porosity. Compressive testing under both room temperature and temperature above the austenite transformation finish temperature are conducted. Porous NiTi specimens perform ductility and show clear superelestic loop with high flow stress and strain.
Two models for predicting the stress-strain curve of porous NiTi under compressive loading are presented in this paper. Porous NiTi shape memory alloy is investigated as a composite composed of solid NiTi as matrix and pores as inclusions. Eshelby’s equivalent inclusion method and Mori-Tanaka’s mean-field theory are employed in both models. In the first model, the geometry of the pores is assumed as sphere. The composite is with close-cells. While in the second model, two geometries of the pores, sphere and ellipsoid, are investigated. The pores are interconnected to each other forming an open-cell microstructure. The two adjacent pores connected along equator ring are investigated as a unit. Two pores interact with each other as they are connected. The average eigenstrain of each unit is obtained by taking the average of each pore’s eigenstrain. The stress-strain curves of porous shape memory alloy with spherical pores and ellipsoidal pores are compared, it is found that the shape of the pores has a nonignorable influence on the mechanical property of the porous NiTi. Comparison of the stress-strain curves of the two models shows that introducing of the average eigenstrains in the second model makes the predictions more agreeable to the experimental results.
Dynamic deformation behavior of TiNi (superelastic grade) and TiNiCu alloy (shape memory grade) were examined using Split Hopkinson Pressure Bar. The stress-strain curves of the TiNi alloy exibits strain rate sensitivity. The flow stress in the plateau region increased with increasing of strain rate logarithmically, and the on-set stress for stress induced martensite also increased slightly. In contrast, the stress-strain curve of the TiNiCu alloy was found to be much less sensitive to strain rate. TEM observations revealed that the microstructure of the dynamically deformed TiNi is similar to that of the sample before dynamic deformation. In contrast, the dynamically deformed TiNiCu has a fine twinned structure than the sample deformed statically.
Analytical constitutive equation for the dynamic deformed TiNi alloy was proposed by addition of the terms concerning the strain rate effect and temperature change due to adiabatic deformation and latent heat of martensitic transformation, the revised constitutive equation was in a good agreement with the experimental results.