Printability and Effective Behavior of Triply Periodic Minimal Surface Nitinol Structures

  • Ali N. A. Alagha

Student thesis: Master's Thesis

Abstract

Shape memory alloys (SMAs) are smart materials that are capable of recovering large inelastic strain by heating or unloading, depending on loading history. Nitinol is the most widely used shape memory alloy, owing to generally superior functional stability, biocompatibility, damping characteristics, and high resistance to oxidation than other SMAs. Nevertheless, Nitinol suffers from poor workability and results in high tool wear during fabrication, which makes it difficult to shape into complex geometries using conventional methods. These issues can be readily addressed through additive manufacturing (AM). In this work, Nitinol-based TPMS structures are investigated through both experimental and numerical-based techniques, using additive manufacturing (AM) and finite element analysis (FEA). Compared to other lattice structures, TPMS lattices generally show better mechanical performance, fluid permeability, and energy absorption and dissipation. In the experimental part, only the bottom layers of the samples were successfully built. Printability analysis of the fabricated samples showed that substrate material, pre-heat temperature, and the type of re-coater blade are the main reasons behind additive manufacturing process stoppage. Furthermore, FEA and numerical homogenization considering TPMS primitive, IW-P, gyroid, and diamond unit cells subjected to different loading conditions were investigated. Under uniaxial loading, the results showed a slight increase in effective properties with temperature and a dramatic increase with rising the relative density of the TPMS unit cells. A comparison study among the four types of TPMS structures showed that diamond has superior effective stress evolution than others. Under biaxial loading, the onset and subsequent thresholds of phase transformation were determined. At lower relative density, the initial phase transformation loading surfaces corresponding to the different geometries considered were found to be reasonably well represented by an anisotropic Hill's and Misses criterions. The observed fit, which also depends on geometry, degenerated for Misses model as the effective martensite volume fraction increased, while properly captured by Hill's model. The determination of subsequent loading surfaces as a function of the effective volume fraction of martensite showed a nonlinear hardening behavior, which seems to follow a similar trend for the different geometries considered. Ultimately, the loading surfaces were found to reach an asymptotic state with distinctly different features compared to their initial shapes.
Date of AwardJun 2021
Original languageAmerican English

Keywords

  • Shape memory alloys
  • Triply periodic minimal surfaces
  • Additive manufacturing
  • Smart materials
  • Effective behavior
  • Transformation temperature

Cite this

'