Mark Nave

Reliable Titanium Aluminide Texture Measurement Using EBSD

γ-TiAl has a tetragonal structure that is almost cubic (c/a ≈ 1.02). As a result of this, the electron backscatter diffraction patterns (EBSPs) of orientations related by 90°<100] rotations are almost identical. Reliable orientation measurement during automated EBSD mapping requires distinguishing very small differences in the positions of certain bands within the EBSPs. EBSD was used to measure the micro-texture and bulk texture of γ-TiAl hot-deformed under conditions approximating plane-strain compression. The reliability and accuracy of the indexing was assessed by inspecting the orientation maps for misindexing and by comparing the texture measured using EBSD with that measured using X-ray diffraction (XRD). Using Advanced Fit refinement and high-resolution EBSPs produced reliable and accurate results. Substantial misindexing was present when the EBSPs were indexed using the Hough transform method alone.

On the Production of Randomly-Oriented Recrystallization Nuclei for Selective Growth Experiments

The ideal starting condition for selective growth experiments is one having a layer of randomly-oriented nuclei adjacent to a matrix with negligible orientational variation but sufficient stored energy to promote growth. In practice, cutting or deformation processes are used in an attempt to approximate these ideal conditions, but the degree to which this is achieved has not been rigorously quantified. In this work, Fe-3wt%Si single crystals were cut or deformed using six different processes. The variation in texture with distance from the cut or deformed surface was measured using electron backscatter diffraction (EBSD) in a field emission gun scanning electron microscope (FEG-SEM) in order to assess the ability of each process to create conditions suitable for selective growth experiments. While grooving with a machine tool produced the best spread of orientations at the cut surface, the suitability of this process is diminished by the presence of a differently-textured deformed layer between the cut surface and the single crystal matrix. Grinding produced a less ideal distribution of orientations at the cut surface, but the presence of these orientations in a very thin layer adjacent to the matrix makes this process preferable for preparing crystals for selective growth experiments, provided the results are corrected for the deviation in the distribution of nuclei orientations from a random distribution.

Development of Boundary Misorientations During Grain Growth in Silicon Steels

Changes in grain size, texture and misorientation distributions have been monitored during extensive normal grain growth in 3%Si steels. The boundary misorientation distributions deviate significantly from the Mackenzie relationship. Comparisons of correlated and uncorrelated distributions show large excesses of low angle boundaries. However, these are not a result of low energy boundaries being favoured during grain growth since the deviation diminishes as growth proceeds. The effect originates in the nucleation of grains in colonies of similar orientation during primary recrystallisation. A slight tendency for promotion of 60º boundaries may indicate some preference for the retention of lower energy twin boundaries during grain growth in silicon steel.

Evaluation of Selective Growth in a Fe-2.8%Si Single Crystal Using Rodrigues-Frank Space

A Fe-2.8%Si single crystal was scratched in order to randomise the texture in the neighbourhood of the notch. Annealing resulted in recrystallization and grain growth starting from the deformed zone. Misorientations between the single crystal matrix and the grown grains were gathered and were studied in order to investigate the possibility for selective growth based on a specific misorientation. However, instead of studying the misorientation angle or axis profiles separately in a 1D or 2D projection a full misorientation analysis was carried out in the 3-dimensional Rodrigues-Frank misorientation space, which offers an unambiguous interpretation of the data because no features are hidden or masked by a projection. It is concluded that the selective growth phenomenon following the <110>26.5deg misorientation relationship is strongly supported by the gathered orientation data, after appropriately normalizing these data with respect to a random misorientation distribution.