SUMMARY

Many advanced alloys exhibit hierarchical microstructures, where important features at different length scales contribute to the overall properties of the material. However, the evaluation of material mechanical properties and microstructure statistics at multiple length scales present formidable challenges. First, most of current methods used for mechanical testing at the different material structure/length scales require substantial investment of time and effort to produce very limited amount of data. The second major obstacle comes from the need to segment the features of interest in the raw microstructure images for evaluation of microstructure statistics. One of the major challenges in this process is that the successful and accurate segmentation is highly dependent on the user’s expertise in image processing functions. These two challenges have been a significant hurdle in collecting sufficiently large experimental datasets of microstructures and their properties necessary for materials innovation efforts, such as advancement/refinement of physics-based composite theories. This work aims to bridge these gaps by developing and demonstrating protocols for reliable and high-throughput experimental evaluation of microstructures and their mechanical properties at multiple resolutions. First, the recent advances in spherical indentation stress-strain protocols are extended and applied to multiresolution testing. These protocols are employed for evaluation of mechanical properties at the individual microscale constituent level as well as at the macroscale. Second, to advance the reliability and accuracy of microstructure characterization, an image segmentation framework is developed to achieve systematic image segmentation on broad classes of microstructures utilizing widely available image processing tools. This work develops and validates these new protocols in an investigation into the mechanical response of thermally aged ferrite-pearlite steel samples.