Analytical and Experimental Nanomechanical Approaches to Understanding the Ductile-to-Brittle Transition

Abstract

This dissertation presents progress towards understanding the ductile-to-brittle transition (DBT) using a mixture of nanomechanical experiments and an analytical model. The key concept is dislocation shielding of crack tips, which is occurs due to a dislocation back stress. In order to properly evaluate the role of these interactions, in-situ experiments are ideal by reducing the number of interacting dislocations and allowing direct observation of cracking behavior and the dislocations themselves. First, in-situ transmission electron microscope (TEM) compression experiments of plasma-synthesized silicon nanocubes (NCs) are presented which shows plastic strains greater than 50% in a semi-brittle material. The mechanical properties are discussed and plasticity mechanisms are identified using post-mortem imaging with a combination of dark field and high-resolution imaging. This observations help to develop a back stress model which is used to fit the hardening regime. This represents the first study of its kind where back stresses are used in a discrete manner to match hardening rates. However, the important measurable quantities for evaluating the DBT include fracture toughness values and energetic activation parameters for cracking and plasticity. In order to do this, a new method for doing in-situ fracture experiments is explored. This method is pre-notched three point bending experiments, which were fabricated by focused ion beam (FIB) milling. Two different materials are evaluated: a model ductile material, Nitronic 50, an austenitic steel alloy, and a model brittle material, silicon. These experiments are performed in-situ scanning electron microscope (SEM) and TEM and explore different aspects including electron backscatter diffraction (EBSD) to track deformation in SEM scale experiments, pre-notching using a converged TEM beam to produce sharper notches better replicating natural cracks, etching procedures to reduce residual FIB damage and elevated temperature experiments. Lastly, an analytical method to predict DBTs is presented which can account for effects of strain rate, temperature and impurity presence. The model is tested by pre-existing data on macroscopic compact tension specimens of single crystal Fe-3%Si. Next, application of the model to nano/micro scale fracture toughness experiments is explored and the large number of confounding variables is discussed in detail. A first attempt at fitting is also presented.