The Role of the Boundary Plane in Grain Boundary Engineering

This paper reports new data on the distribution of grain boundary planes in commercially available GBE copper. The results show that nearly 60% of grain boundary length was Σ3, i.e. annealing twin related. There was a slightly enhanced proportion of {111} boundary planes, even after Σ3 boundary segments were excluded from the data set. There was a high proportion of boundary plane length on the 110 zone, which corresponds to asymmetrical tilt boundaries misoriented on the <110> axis. Some of these were geometrically necessary Σ9 and Σ27 boundaries. The findings support the view that grain boundary plane engineering is a viable way forward. INTRODUCTION: ‘Grain boundary engineering’ (GBE) refers to the manipulation of microstructure, usually via iterative thermomechanical processing, in order to improve material properties. These properties can be interface transport related phenomena such as intergranular corrosion or cracking, or overall properties such as ductility. A recently published set of papers has overviewed the current status of grain boundary engineering (e.g. Homer et al [2006]). Almost always GBE relies on prolific multiple annealing twinning. In coincidence site lattice (CSL) nomenclature a twin is a Σ3 boundary. A Σ9 boundary (so-called ‘second order twin’) arises from the conjoining of two Σ3s at a triple junction. Many GBE investigations rely on a CSL approach to categorise and explain the observations. However there is growing opinion that the CSL approach is inadequate because it is a misorientation-based characterisation, and the properties of high angle grain boundaries relate instead to the crystallography of the grain boundary plane (Randle [2006]). Recently, populations of grain boundary planes can be measured via an extension to the electron backscatter diffraction (EBSD) orientation mapping technique in a scanning electron microscope (SEM). This is known as the ‘five parameter analysis’ and is described in detail elsewhere (Saylor et al [2004]). The new analysis has allowed the emphasis when measuring grain boundary crystallography to be focussed on the distribution of interface planes rather than on the misorientation alone. Other recent methods also exist to extract the boundary plane (Homer et al [2006]). In this paper, we report new data on the distribution of grain boundary planes in GBE copper. The findings support the view that grain boundary plane engineering is a viable way forward.