Considering now the in-planeconnectivity along thedirection at the perpendicular to the printing direction (Fig. 12c, d), a few bridges can be found between two channels in the [−45°/45°] case and much more are found in the [0°] case. This indicates that alternating the printing of a layer with another one at [90°] does not only enable to reduce the mean volume fraction of pores but also considerably reduces the con- nectivity of pores in the direction perpendicular to the printing. Fig. 12e, f enables to confirm the propensity of crisscross printing to reduce pores network connectivity: they show that out-of-plane bridges exist betweentwo in-plane porosity layers intheunidirectional printing case but none can be seen in the crisscross printing case. Focusing on the unidirectional case, the strong anisotropy of the porosity network can be thought to be responsible for the strong dif- ferences in the tensile behavior of the [0°] tensile specimens and the [90°] ones. But it cannot be the only source of differences: although [90°] tensile specimens figure out ahigher volume fraction of pores and higher pore connectivity distances (in the three principal directions ofthe specimen) than crisscross specimens, they display tensile properties which are close to those of the crisscross specimens.