It sounds like you might be talking about dislocations rather than grain boundaries; two models of precipitate strengthening involve dislocations cutting through vs. looping around a particle. It this possibly the case?
Actually, I would like to ask about the dislocation case too. What I originally meant was, could what occurs to dislocations also occur for grain boundaries? I'm not very clear on the concepts.
Suggest looking at Grant et al., "Grain-boundary migration in single-phase and particle-containing materials," J Mater Sci (1984) and Randle et al., "Interactions of grain boundaries with coherent precipitates during grain growth," Acta Metall (1986).
there are two kinds of fracture: one is through the grain, and the other is along the grain. usually, the grain boundary is the place where defects get together, so it is the most un-stable. If in a ceramic, fracture will be the grain boundary