Shape memory alloys are materials with remarkable properties that stem from a martensitic transformation. One of the fundamental aspects of shape memory and pseudoelastic behavior that is not well understood is how the matrix accommodates the large strain associated with the transformation. Theoretically, accommodation may be achieved either by matrix plasticity or by inducing other transformation variants. This three-year GOALI proposal will develop new, fundamental insight into the pseudoelastic shape memory behavior of several Ni-Ti based alloys, including a ternary Ni-Cu-Ti, in collaboration with GM Research. The focus will be on alloys that undergo different martensite transformations and exhibit disparate functional fatigue properties. The approach is to meld innovative micron-scale mechanical tests and advanced microstructural characterization with analytic and novel microstructure-based modeling. Uniaxial deformation of focused-ion-beam-machined microcrystals will be used to probe the static and cyclic response as a function of matrix crystal orientation, and in order to directly measure the mechanical response (stress, strain, work output) for individual variants. Post-mortem characterization using transmission electron microscopy of the remnant substructure will be coupled with analytic modeling of the possible transformations. This approach will also yield directly the relative conditions for martensite transformation versus matrix plasticity over a range of component sizes. A new microstructure-based finite element approach will be developed that explicitly tracks local, discrete phase transformations coupled with rate-dependent crystal plasticity. For the first time, this will enable the treatment of size effects and the generation of local plasticity associated with transformations?a crucial step to understand and enhance pseudoelastic and shape memory response. The proposed effort integrates advanced characterization at The Ohio State University and the GM Research center with new experimental techniques at two DOE labs ? high temperature nanoindentation/pillar testing at Oak Ridge National Laboratory and in situ pillar testing at the National Center for Electron Microscopy.
NON-TECHNICAL SUMMARY: Shape memory alloys (SMAs) are materials with remarkable properties that include the ability to bend and stretch to large extent under an applied load, then spring back to their original shape when the load is removed. In addition, SMAs can change shape when heated, after which they may either maintain their newly acquired shape or return to their original shape after cooling back down to room temperature. The automotive industry has recognized the phenomenal potential of SMAs since they are remarkably simple actuation devices compared with conventional motorized actuators. For instance, they could be used as small, ?solid-state? motors that could be used to reconfigure a wide range of components, greatly reducing the complexity of such systems. However, commercially available SMA materials are not presently operated at their full potential due to degradation in their shape-changing capabilities after experiencing many temperature or stress cycles (?functional fatigue?). This program is designed to develop a fundamental understanding of the materials science aspects associated with this degradation process, as well as the development of modeling capabilities to predict and improve the functional fatigue performance for automotive, medical and other applications. A vigorous program of interchange between OSU and GM will stimulate efficient transfer of knowledge and will provide ample mechanisms for experiencing both academic and industrial environments. In an exciting outreach effort, we will develop a high school inquiry-based teaching module about shape memory alloys and their applications. Several mechanisms insuring insertion of this module into local high schools have also been defined.
The three-year GOALI project has enabled several important advances in the characterization and simulation of shape memory alloys. These alloys can be used as solid state actuators and shape changing devices in numerous application areas, including transportation and medicine. A primary goal of this work is to understand the microstructural evolution and resultant functional fatigue behavior of NiTi-based shape memory alloys. At present, commercial application of these materials is limited by degradation of their actuation and shape changing properties during repeated use. Advances in characterization include the development and optimization of micron-scale pillar compression testing capabilities. This effort has enabled study of the performance of microcrystalline regions within large polycrystalline samples. Tests at ambient and high temperature tests are now possible through acquisition of a state-of-the-art nanoindentation system. This system can be operated in a scanning electron microscope (SEM), allowing for detailed images of phase evolution during compression testing. These novel mechanical tests are paired with scanning transmission electron microscopy (STEM) analysis of the microcrystals after testing. Fundamental stress-induced martensite transformation events have been isolated and shown to trigger crystalline slip in the vicinity of these events. Such slip events are associated with dimensional instability of shape memory allows during repeated actuation and the aforementioned degradation in of properties. Advances in simulations include the development of micromechanics-based finite element modeling and Fourier transformed based micromechanics modeling. The former has been developed using a commercial finite element code (ABAQUS) and the later uses custom coding (in MATLAB). These codes capture the essential features of martensitic transformations. This is achieved by incorporating the phenomenological theory of martensite transformation. The finite element simulations can capture the response of both single-crystal and polycrystalline material. They can predict the evolution of the most favorable martensite plate(s) to form during the aforementioned compression testing and also predict plasticity due to intergranular constraints. The MATLAB-based simulations predict the most stressed slip system in the vicinity of martensitic plates at much smaller scales than the finite element simulations. These experiments and simulations were the first of their kind to show a direct link between the martensite transformation behavior of shape memory alloys and the accumulation of crystal defects. Further testing has verified both the experimental and computational efforts for a variety of single crystal orientations. STEM observations of a variety of different thermomechanical test conditions have been made. An in situ STEM heating study has revealed austenite defect substructures in material compositions that are fully martensitic at room temperature. Modeling efforts are currently underway to provide a direct comparison between these results and the micropillar observations. The MATLAB simulations described above have been extended and applied to thin foil samples that are typically used in STEM studies. Martensite plates most likely to form under tensile loading have been identified. Volume fractions of the correspondence variants forming the most efficient martensite plate are predicted. They match well with the TEM observation of the thin foil. An important conclusion of this modeling exercise is that the martensite plate types that form in thin foils are different and cannot be predicted by the traditional phenomenological theory of martensite for bulk materials. Finite element simulations have been used to systematically study how the shape memory response of individual grains within larger polycrystals varies from grain to grain. The crystallographic orientation of the grains, grain neighborhood, polycrystalline texture and deformation history of the sample are identified as the principal factors affecting axial transformation strain at the local (grain) scale. A study is nearing completion to predict bimodal grain textures to optimize shape memory performance. This effort will guide texture development of shape memory alloy wires that are used in a variety of applications. Work is underway to develop and implement a new microstructural FE model based on the phase field (PF) method. A primary goal is to simulate phase transformations and reorientation at the scale of correspondence variants. Experimental results suggest that the fundamental mechanisms of degradation and dimensional instability occur at this scale. The driving force for phase transformation and reorientation consists of a polynomial-based Landau free energy functional and gradient based interfacial energy term. Compared to PF formalisms that are solved using transform techniques, the FE approach can handle arbitrary boundary conditions and model large deformation/rotation associated with phase transformations. It will enable study of the underlying degradation mechanisms in shape memory alloys at nm scales and couple closely with the electron microscopy techniques. Industrial efforts with NASA Glenn Research Center and General Motors Research and Development as well as international collaborations with Ruhr University (Bochum, Germany) have been facilitated by this funding opportunity. Research findings have been co-published with these partners and will continue in future work.
