Performance of UV-molded versus injection molded maskless microlens arrays for multi-aperture projection

Maskless microlens arrays (MLA) for multi-aperture projection offer high transmission due to absence of buried absorbing mask structures for shaping the pattern. Arbitrary shaped lenslets arranged in a high fill-factor array form the entrance MLA, the apertures of each entrance lenslet are projected towards the far-field by regularly shaped lenslets of the exit array. Resulting tandem MLA is arranged in a fly's eye homogenizer (FEH) configuration. Such MLAs are mastered by grayscale photolithography and replicated as UV-molded polymer-on-glass (POG) elements. Furthermore, such maskless projectors can be realized using alternative replication technologies, primarily - injection molding (IM), which offers low cost for large area replication. MLA designed for POG replication can be adapted for IM replication by adjustment of certain parameters, e.g., thickness due to material substitution. Since the shape of the lenslets remain same for both processes, the same MLA masters can be used to generate the tooling molds necessary for IM replication. However, unlike POG replication, IM technology lacks active alignment of the entrance and exit arrays which makes IM replicated MLAs more vulnerable to crosstalk and stray light. Hence, a trade-off between ease of manufacturing (low costs) and projected image quality (sharpness, contrast, stray light) should be considered when dealing with IM replication. In this work we describe an exemplary IM replicated maskless MLA, based on masters of a previously realized POG replicated MLA for an automotive projected blinker. We discuss the design adaptation to IM and present profilometric and photometric characterization of the IM replicated MLA. We also compare the characterization results of the IM MLA samples with that of 'gold-standard' POG replicated MLA and discuss performance, quality of projection and limitations of IM technology.