Ductile fracture of metals involves a sequence of overlapping processes that include the nucleation, growth and coalescence of voids. Of these, the last is the most important because it dictates the ductility of metals but is however the less understood. As the coalescence is a stochastic event occurring over very short strains, it is really difficult to capture experimentally. Attempts to fabricate model materials that would simplify the analysis of the ductile fracture process have already been made by Babout et al. [1], Gammage et al. [2], Magnusen et al. [3], Jia and Povirk [4] and Nagaki et al. [5]. However, they are of limited help when one wants to study the coalescence event in detail because of their complicated microstructures, of there limited ductility, of the controlling effect of the nucleation event, of the lack of constraint (2D approaches instead of 3D) or simply because they do not reproduce the key features of the microstructure (like the void size for instance).