Grain Boundary Character Distributions In Isostructural Materials.

摘要

Anisotropic grain boundary character distributions (GBCDs), which influence macroscopic materials properties, are thought to be controlled by the grain boundary energy anisotropy. Structurally, grain boundary could be viewed as two free surfaces joined together. Grain boundary energy could be simply defined by the total excess energy for creating two free surfaces minus the energy gained when new bonds are formed between these surfaces. This implies that different crystal structure should have different GBEDs and GBCDs. It was recently discovered that grain boundary energy distributions (GBED) in isostructural materials, a class of materials that share the same crystal structure, are directly related to one another. This suggests that GBCDs in isostructural materials might also be related in a similar way. To test this hypothesis, electron backscatter diffraction (EBSD) was used to map grain orientations in Ag, Au, Cu, Fe, and Mo. The GBCDs were determined from the stereological interpretation of EBSD maps containing on the order of 100,000 grains. It was found that the GBCDs of face-centered cubic (FCC) metals are statistically correlated, while the GBCDs of body-centered cubic (BCC) Fe and Mo are not correlated to the GBCD of FCC metals. The degree of the correlations among the FCC metals is weaker if there are significant differences in grain shape or texture. For example, Ag has the weakest correlation to the other FCC materials and also has quantitatively different grain shapes and texture. The relationship between the populations and energies of grain boundaries was also studied. By comparing the GBCDs of Al, Au, Cu, and Ni to the energies of 388 grain boundaries previously calculated by the Embedded Atom Method (EAM), we observed a moderately inverse correlation between the relative areas of grain boundaries and their energies. Interestingly, there are strong inverse correlations between the energies and populations of the most common grain boundaries (Sigma3, Sigma9, and Sigma27). Because the enhancement of twin related boundaries due to the prevalence Sigma3 boundaries results in a decrease in the grain boundary populations for the other boundary types, this inverse correlation is influenced by the crystallographic constraints at triple junctions. In other words, having an anisotropic misorientation distribution with strong maxima for certain boundaries biases the inverse correlation between grain boundary population and energy for other boundaries and causes different slopes at each misorientation. Interestingly, the inverse correlation at each misorientation is consistent with the Boltzmann distribution. Based on our results, it is possible to predict the GBCDs and GBEDs in isostructural polycrystalline materials by using a single GBCD and GBED. This principle is demonstrated by predicting the GBCD and GBED of Actinium (Ac).;In this thesis, the results demonstrated that the GBCDs of isostructural materials are correlated with one another and the magnitudes of correlation coefficients varied. Reduced correlations were observed when there were differences in the microstructure and crystallographic texture. The inverse relationship between grain boundary population and energy is more strongly correlated at each misorientation than over the entire five macroscopic parameters of grain boundary, especially when there is significant misorientation texture. This relationship leads to GBCDs of isostructural materials that are also more strongly correlated at each misorientation than over the entire grain boundary space.;To investigate the GBED in the isostructural BCC metals, the energies of 408 grain boundaries in Fe and Mo were computed using atomistic simulations based on the embedded-atom method (EAM) potential. We found that the calculated boundary energies in Fe and Mo were strongly correlated and scaled with the ratio of the cohesive energy divided by the square of the lattice constant (Ecoh/a02). We would expect that the GBCD of Fe and Mo might be correlated in a similar manner to that of FCC metals. To test this hypothesis, we compared the GBCDs of Fe and Mo. We found that the GBCDs of Fe and Mo are moderately and strongly correlated when all boundary types and only Sigma3 boundaries were considered, respectively.