Analysis of grain boundary networks and their evolution during grain boundary engineering

The goal of grain boundary engineering is to increase the fraction of so-called special grain boundaries, while decreasing the contiguity of the remaining random boundaries which are susceptible to intergranular degradation such as cracking, cavitation, corrosion and rapid self-diffusion. In the present work, we describe a technique for the quantitative experimental study of grain boundary network topology, with an emphasis on the connectivity of special and random grain boundaries. Interconnected grain boundary networks, or "clusters", of either entirely random or entirely special boundaries are extracted from electron backscatter diffraction data on a Ni-base alloy, and characterized according to their total normalized length (their "mass"), as well as their characteristic linear dimensions. The process of grain boundary engineering, involving cycles of straining and annealing, is found to substantially reduce the mass and size of random boundary clusters. Furthermore, quantitative assessment of the boundary network topology shows that the special grain boundary fraction is a poor predictor of network topology, but that the higher-order correlation derived from triple junction distributions can successfully describe the length scales of random boundary clusters. (C) 2002 Acta Materialia Inc. Published by Elsevier Science Ltd. All rights reserved.