Influences of processing and composition on the grain boundary character distribution.

다국어 초록 (Multilingual Abstract)

Experiments were performed to observe the effects of processing and composition on the grain boundary character distribution (GBCD) of SrTiO 3, MgO, and Ni. Anisotropic interfacial energies and, in the case of nickel, crystallographic constraints, are expected to influence the evolution of the GBCD during processing.
Interrupted grain growth experiments were performed with SrTiO3 to observe the evolution of the GBCD with grain growth. The GBCD was determined at three time steps during grain growth and quantitative comparisons were made. Guidelines for quantitative comparisons of grain boundary plane distributions (GBPDs) and GBCDs were established. Relative grain boundary energies were evaluated using two independent techniques to observe the relationship between grain boundary population and energy and the role that grain boundary energy has in determining the GBCD.
It was found that anisotropic interfacial energies dictate an anisotropic GBCD which develops very early in the grain growth process for SrTiO 3. The GBPD remains constant even as the grain size increases. Relatively low energy {100} grain boundary planes are favored. The five parameter GBCD evolves during grain growth to eliminate relatively higher energy grain boundaries while lower energy grain boundaries are preserved.
Doping was explored as a processing technique to alter the GBCD in MgO. The GBCDs of 3000 ppm Ca-doped MgO and undoped MgO were measured and quantitatively compared. Relative grain boundary energy anisotropy was determined for each material and compared to the anisotropy of the GBPD.
Calcium doping in MgO increases the anisotropy of the GBPD and GBCD. Specifically, the relative population of {100} grain boundary orientations increases with doping. In the five parameter GBCD of the pure material, some misorientations favor {100} boundary plane orientations and in these cases, the doping increases this preference. For misorienations in the pure material that favor {111} boundary plane orientations, the preference for {111} is diminished or the preference switches to {100} orientations. The anisotropy of the grain boundary energy distribution increases with doping.
Iterative thermomechanical processing was explored as a technique to alter the GBCD in nickel. GBCDs were determined before and after iterative thermomechanical processing for both commercially available "grain boundary engineered" (Integran) and high purity nickel. Triple junction analysis was performed in which triple junctions were classified according to the types of grain boundaries that formed the junction. The relative grain boundary energy distribution was also determined for high purity nickel.
Grain boundary energy was also found to be a factor of influencing GBCD evolution during iterative thermomechanical processing. Boundaries with relatively higher energy decrease in population while lower energy boundaries increase in population. Specifically, random boundaries are preferentially eliminated while the population of Sigma3 boundaries increases. However, because the evolution mechanism is different than in grain growth, and relies more on local rearrangements of boundaries rather than large scale elimination of interfacial area, constraints placed on the resulting network of boundaries influence the GBCD as well. Thus, interfacial energy is not the sole factor that determines the GBCD in nickel. The population of incoherent Sigma3 boundaries increases more than the population of the much lower energy coherent Sigma3 boundaries. Sigma9 and Sigma27 boundaries are also observed to increase in population. These boundaries tend to be the result of interacting Sigma3 boundaries. This conclusion is supported by the triple junction populations that show an increase in triple junctions that are comprised of two or three Sigma3 n type boundaries, where n = 1, 2, or 3. A mechanism is proposed that accounts for the increased population of incoherent Sigma3 boundaries, the increased populations of Sigma9 boundaries, and the observed triple junction populations.