GBE is an emerging field, which promises methods to improve the performance of materials, whose deg­radation in service is caused by the presence of high angle boundaries. The concept, first proposed59 by Prof. T Watanabe in the early 1980s, envisages improvement ofproperties ofmaterials by controlling the grain boundary character distribution (GBCD). Many processes like diffusion, precipitation, segrega­tion, sliding, cavitation, and corrosion are kinetically faster along high angle grain boundaries. Hence, it is possible to decelerate these detrimental pro­cesses by replacing the random boundaries with low energy ones, coincident site lattice (CSL) boundaries (denoted by the ‘sigma number,’ S, which is defined as the reciprocal of the fraction of lattice points in the boundaries that coincide between the two adjoining grains on the basis of CSL model). Another prerequi­site for GBE is to completely destroy the interconnec­tivity of random grain boundary network. The insight in the field of GBE was achieved with the advent of computer assisted EBSD (electron back scatter diffraction) technique developed during the 1980s.

The embrittlement in ferritic steels is known to be caused by segregation phenomena. The kinetics of segregation can be controlled by suitable selection of the nature of grain boundaries. GBE has been applied60-63 to combat embrittlement problems in ferritic steels. The task of carrying out GBE using experimental methods is time consuming. Hence, it is prudent to resort to computational methods, which need to be validated using selected experiments. A 3D Poisson-Voronoi grain structure, simulated using MC technique was employed to study60 (Figure 12(a)) intergranular crack percolation using percolation theory. The percolation threshold was estimated to be 80%. To apply this model to specific alloys like ferritic steel, system specific characteris­tics need to be incorporated61 in the model. One such attempt is to define the propensity of the grain boundaries for propagation of cracks based on rela­tive values of the grain boundary energy and the

0.185 0.180 0.175 0.170 0.165 0.160 0.155 0.150 0.145 0.140 0.135 0.130 0.125

0 5 10 15 20 25 30 35 40 45

(d) Charpy impact energy (J)

energy required for propagation of cracks. These calculations were carried out (Figure 12(b)) for two different grain sizes. The prediction of finer grain size being favorable to reduce embrittlement was con­firmed (Figure 12(c)) experimentally. The GBCD, that is, the distribution of various grain boundary types has been evaluated62 in modified 9Cr-1Mo fer­ritic steel using EBSD technique. The experimental observations confirmed the reduction of DBTT by 20 K with reduction in grain size. The fractal analysis of the fracture surface demonstrated (Figure 12(d)) that the tortuous path which cracks need to follow in fine grain sample is responsible63 for the observed reduction in the propensity for embrittlement.

It is shown clearly that the low energy boundaries can be introduced in engineering materials in three different methods: preferential nucleation of low angle boundaries around twins or controlled recov­ery or orientation relations during phase transforma­tion, if some of the variants happen to result in CSL boundaries. Significant improvements in properties using GBE have been achieved64 in many austenitic stainless steels, in contrast to ferritic steels. The major challenges in the application of GBE to Cr-Mo fer­ritic steels arise from the following factors: lower twinning probability, higher stacking fault energy, and limited variants with CSL boundaries during g! a transformation during cooling.