This project is based on collaboration between Harvey Mudd College (HMC) and the School of Materials Science and Engineering and the Electron Microscope Unit at the University of New South Wales (UNSW) in Australia. It is also a Research in Undergraduate Institutions activity, as the award supports international materials science research experiences for HMC undergraduate students. The project focuses on resolving a longstanding controversy about the nature of arrays of aligned dislocation boundaries that are generated during rolling of deformed face-centered cubic and body-centered cubic metals with intermediate to high stacking fault energies. The structure and origin of these dislocation boundaries are of substantial interest due to their role in determining anisotropy in yield stress and strain hardening. Insight into the true character of these microscale structures is essential for advancing the predictive capabilities of physically-based models for macroscale mechanical properties. The two opposing theories of the structure and origin of these aligned boundaries in deformed polycrystals are: (i) they are oriented along certain crystallographic planes, or (ii) their alignment is dictated primarily by the macroscopic stress state during plastic deformation. Current evidence supporting these theories is based on two-dimensional data that does not necessarily reveal the true nature of deformation microstructures and can lead to erroneous interpretation. To definitively resolve the issue and to further modeling capabilities, three-dimensional (3D) orientations of the boundaries are required. Data in 3D is collected using focused ion beam-electron backscatter diffraction (FIB-EBSD) tomography. EBSD maps of FIB-generated serial sections are combined in post-processing to generate full crystallographic volumes capable of revealing many types of structural features at submicron resolution.
HMC students lead the development of new computational tools required to efficiently and accurately analyze the large 3D data sets that result from this method. They also refine FIB-EBSD methods and present their findings in peer-reviewed journals and national or international conferences. At UNSW, the supported undergraduates and the principal investigator have extensive access to electron microscopy facilities and training unavailable at HMC. The project therefore provides an extraordinary opportunity for collaboration between UNSW's experts in physical metallurgy and electron microscopy and U.S.-based undergraduate students. This award is co-funded by the Division of Materials Research and the Office of International Science and Engineering.
The complex deformation and annealing procedures associated with thermomechanical processing affect the microstructure and texture of most metal components. These, in turn, have an influence on the final properties of the fabricated materials and dictate their use in a wide range of applications including, for example, the superstructure and fabricated panels of every automobile, ship and airplane. Therefore it is important that the properties of deformed and recrystallized metals are optimized. This can be achieved by a better understanding of the nature of the microscale properties of metals in their deformed state. This project addressed a longstanding and controversial issue concerning the nature of specific deformation structures, called microbands, that are generated in rolling and other plastic deformation of commercially significant metals. The key result of this work was significant evidence, in multiple materials and with multiple modes of deformation, that both of the previous theories have merit. In each sample we found evidence of microband boundaries that were oriented both according to the crystal structure of the metals and also according to the loading they were subjected to. The fact that both orientations can be seen in close proximity has implications for understanding of material properties and for future modeling efforts of the dynamic process of plastic deformation. Figure 1 shows reconstructed surfaces of microbands within the same crystal grain of a commercially pure aluminum sample. That the orientation of the left surface is dictated by the crystal structure is shown by the correspondence of the grey surface normal dots with the red crystal plane marker. The correspondence of the right surface with the stress generated during plastic deformation is shown by the location of the grey surface normal dots with the blue stress plane marker. In conducting experiments, we mapped the 3D crystal orientation of volumes of the metal using electron backscatter diffraction in an electron microscope. While working toward the main scientific question, we also advanced methods for this technique. We also developed signifcant computational methods for segmenting the data sets into separate microbands, which is needed to study the microband boundaries of interest. The most powerful method, a modification of the existing fast multiscale clustering (FMC) method, can be used not only to segment microband data sets that contain gradual and subtle boundaries but also data sets with distinct boundaries as shown in 2D examples in Figure 2. Although the novel FMC implementation was developed for the challenging microband data sets, it has wider applicability and is being incorporated into an open-source software package for working with this kind of data. Through this grant and additional resources leveraged with it, nine undergraduate students had significant research overseas research experience in materials science at the University of New South Wales in Australia. All gained experience in experimental and computational methods and in writing papers for publication. In each of the three funded summers one of the two students directly supported by the grant was female, and one student was African American. Two had never been out of the US before this experience. Most of the students have begun or are applying to PhD programs.
