Elucidating deformation mechanisms in shape memory alloys using 3D X-ray diffraction

Abstract

Advanced materials are internationally recognized as a foundation for new capabilities, tools, and technologies that meet urgent societal needs including clean energy, human welfare, and national security [1]. They are also recognized to be very costly to develop and slow to become commercially available, often requiring decades of research before entering the market. “Advanced materials” broadly describes innovative materials that have atypical sizes, microstructures, deformation mechanisms, and/or material responses. These atypical characteristics enable major, previously impossible technological breakthroughs, yet the previously unknown simultaneously presents new challenges to overcome in developing and certifying those technologies. To accelerate the deployment of advanced materials, major advancements in modeling and characterizing advanced materials are needed to accelerate our ability to understand and predict their behaviors. This dissertation presents the need for, challenges of, and solutions to adapting modern X-ray diffraction experiments to elucidate the micromechanics of a subclass of advanced materials called shape memory alloys (SMAs). In particular, the previously established far-field and near-field High-Energy Diffraction Microscopy (ff-HEDM and nf-HEDM) classes of 3D X-Ray Diffraction (3DXRD) techniques and microcomputed tomography (μCT) are advanced to create new abilities to study martensitic transformations and twin reorganization in SMAs. Additionally, a recently developed technique called Dark-Field X-Ray Microscopy (DFXM), which pushes the boundaries on the length scales accessible to non-destructively evaluate crystalline materials from microns to nanometers, is used to study SMAs for the first time. This dissertation begins with an introduction to SMAs and HEDM (Chapter 1), followed by a study that illustrates the need for more complicated micromechanical modeling and associated experimental verification data sets (Chapter 2), and then goes on to report on four X-ray diffraction experiments on SMAs, where each experiment contains one or more novel approaches to experimental planning and data analysis (Chapters 3-6). The first experiment (Chapter 3) discusses the challenges presented by the martensite phase in analyzing 3DXRD data sets for SMAs, which make analyzing the data with the traditional techniques extremely difficult or impossible. In general, these challenges may arise with any multiphase material system, especially where the crystallographic system of one or more of the phases is low symmetry (e.g., monoclinic, orthorhombic, etc.). The technique advancement used to address these challenges is to analyze the data using a forward model algorithmic approach, where the diffraction patterns of virtual microstructures are simulated and compared with the experimental diffraction patterns. In this application, the virtual microstructures are limited to those that are theoretically possible according to the Crystallographic Theory of Martensite (CTM) (also called the Phenomenological Theory of Martensite), and a work flow for the algorithmic approach is presented so that more sophisticated micromechanical models can be implemented in the future. We show that this approach is successful in identifying martensite orientations in three single crystal data sets, even when the single crystals have engineering-grade microstructure features that violate the underlying assumptions of the CTM (i.e., have precipitates, inclusions, elastic strain, subgrains, R-phase, plasticity, and are not infinite plates). We also use this implementation of the forward-model algorithm to show that the application of the widely accepted maximum transformation work criterion needs to be modified for cases where SMAs violate the assumptions of the CTM. The second experiment (Chapter 4) presents a study where we elucidate load-induced twin rearrangement, a reversible deformation mechanism by which materials can accommodate large loads and deformations without damage through reversible rearrangements of crystallographic twins. In this in situ experiment, we use a suite of X-ray measurement probes, nf-HEDM, ff-HEDM, μCT, as well as digital image correlation (DIC). The different types of data collected are then correlated, resulting in a more complete understanding of the micromechanics as well as a variety of ways to convey both quantitative and qualitative information. We also present several data analysis techniques for the first time, including a procedure to measure subgrain-scale lattice rotation and elastic lattice strain correlations, a method to distinguish different types of regions in nf-HEDM reconstructions using confidence thresholding, and the first-ever nf-HEDM reconstruction of a monoclinic material, demonstrating the utility of nf-HEDM even for very low crystal symmetries. We show that a specific sequence of twin rearrangement micromechanisms occurs inside macroscopic deformation bands as they propagate through the microstructure, and we show that the strain localization inside these bands causes the lattice to curve by up to 15°, which has important implications on elastic strain, resolved shear stress, and maximizing the twin rearrangement. In the third experiment (Chapter 5), we use nf-HEDM and μCT to study a phenomenon wherein both low-angle and special high-angle grain boundaries appear inside austenite grains in SMAs as a result of load-biased thermal cycling (LBTC). LBTC is what happens to SMAs in most actuation applications, including several aeronautical applications that are either already on the commercial market or are in development. We use nf-HEDM and μCT to quantify the emergence of low- and high-angle grain boundaries through bulk single crystals during LBTC, providing both boundary angle histograms and 3D EBSD-like reconstructions of the gage volume at high temperature, under two different load biases. The emergence of low-angle grain boundaries is the effect of plasticity developed during the forward and reverse transformation processes. The emergence of high-angle grain boundaries is thought to be an effect of deformation twinning, which causes them to have Σ-boundaries, or coincident site lattice (CSL) boundaries, with the surrounding grain. Both low-angle and high-angle grain boundaries will have a significant effect on the functional fatigue and overall performance of the SMA actuator. In the fourth experiment (Chapter 6), we use DFXM to study intragranular elastic lattice strain and orientation fields in the austenite phase in SMAs during thermal cycling with a spatial resolution of 100 nm. The results indicate that the austenite undergoes a long-range lattice rotation away from the transformation front, and that the transformation front is preceded by a large compressive strain. We also show that the regions of the austenite crystal that transform earliest are those surrounding microstructure features (i.e., precipitates and inclusions) as well as one of the free surfaces, and these regions are consistently the first to transform and reverse transform cycle-to-cycle. This experiment is the first-ever DFXM experiment on SMAs, and this chapter demonstrates the utility of this technique in studying embedded interfaces in SMAs and other advanced materials in situ due to its high spatial, orientation, and strain resolutions. The dissertation is concluded in Chapter 7, where the technique and the microstructure mechanics advancements from each of the preceding chapters are summarized. Chapter 7 also discusses open areas of research with regards to each of the experiments, as well as a general outlook on the subject of using in 3D in situ X-ray techniques to studying advanced material systems.