3D model optimization for 3D printing
The use of digital fabrication (or, 3D printing) has been rapidly expanding in the past couple of years. Virtually any 3D model can be fabricated with high precision and from a wide variety of materials. Since the prices of 3D printers are steadily decreasing, they are becoming more available to the general public. However, there are many open problems. There is no guarantee that the designed 3D model will be structurally sound and it can break after printing. 3D printing is time consuming and requires the primary and support materials which can be expensive. This thesis proposes an optimization framework that seeks to optimize computer-designed 3D models for digital fabrication algorithmically, as the manual repairs are tedious or not applicable to every problem. The primary goal of the proposed framework is to save material by converting the 3D model into a thin shell that is segmented into parts. These parts are tightly packed, lowering the build volume, speeding up the printing process and allowing to print objects larger than the maximum build volume. For the models requiring the additional supporting structures, tree-like supports are generated to further lower the material usage. Structurally weak and unprintable parts of the model are optimized as well. The framework can be used as a whole, or each block can be used for the independent optimization. The system was tested on different types of 3D printers and results were compared to the original unoptimized models and existing algorithmic solutions. The results showed that savings were significant for the most of test models. In average, the material usage can be reduced by 40% with simultaneous 25% savings of the printing time. The optimization time (minutes) is negligible to the printing time (hours).^
Bedrich Benes, Purdue University.
Computer engineering|Computer science