Potentials and Challenges of Generative Design for Additively Manufactured Freeform Optics

Additive manufacturing (AM) offers greater design freedom than conventional processes and is promising for producing freeform optics. However, superimposing inaccuracies and anisotropies caused by layer-by-layer AM processes can only in part be modeled analytically and accounted for in the design process. Therefore, new optics design processes are necessary. In conventional optics design, a nominally ideal system is designed first. The analysis and compensation for tolerances is done in the optimization and validation phase. Because in AM of optics the influence of process deviations and anisotropies is far greater, ideal geometries cannot be found in initial designs. A possibility to address this is a generative design approach, analogous to the design of free-form parts in mechanical components. In this approach, a large number of different solutions are developed in parallel to account for the uncertainty in their final properties. This paper discusses a framework, its main challenges and their solutions for a generative design approach in AM of optics. The goal is to provide a design process that helps to fully exploit the freeform opportunities that AM of optics offers