Tessellated arrays of shapes were also considered as 2-D patterns for prismatic etching because they are simple to lay out. Though seemingly absent from the literature, tessellated squares, shown in Figs. 9(b) and 9(c), appeared desirable for this application because they conform to the symmetry of the square micromirror geometry. Hexagonal patterning, such as that used by Su et al.,10 and equilateral triangular patterning are inherently asymmetric patterns and thus will introduce some form of axial asymmetry in the curvature for this geometry. For completeness, implementations of these [see Figs. 9(e)–9(f) and 9(h)–9(i)] were nevertheless examined in part because of the prior work that suggested they may be desirable. The parameterization for each of these models used the shape sidelength dimensions (, , and , respectively), the etch depth and size of the tessellation array (the triangular and hexagonal used a single side as reference due to the asymmetry) as the dependent variables. Relations appropriate for each geometry, similar to Eq. (7), were then applied to space the tessellated elements evenly along the mirror length, with an equivalent spacing near the periphery of the mirror. Element spacing was given an artificially imposed lower resolution limit of to ensure high confidence in fabrication. In all cases, the area in the center of the tessellated pattern was reserved for the post (4 squares, 10 triangles, and a single center hexagonal element) based upon shape size. Thus, the lower tessellated patterns had reserved central areas that were in excess of the square size needed for the baseline geometry, making the higher patterns more favorable from a mass reduction standpoint. Tailored design of these excess areas near the post, as well as the unpatterned edge areas of the asymmetric tessellations, perhaps with a second-higher resolution tessellated pattern, is possible but was outside the scope of this work.