SUMMARY

In this work, I investigate the underlying chemical behavior of grayscale digital light processing (g-DLP) and how the physical output is influenced by factors such as printing conditions and liquid resin composition. I use this understanding of the process to create significant improvements in geometric accuracy using both high level simulations and machine learning techniques. Finally, I employ g-DLP printing in different applications where the heterogeneous printed parts can outperform similar homogeneous isotropic structures. This work improves the accuracy of and enlarges the design space for DLP printing.