DSurf: Scalable Computational Methods for 3D Printing Surfaces
Italy has possibly the most respected design community worldwide. "Made In Italy" products are known for their quality, functionality and elegance. But as a country, Italy is not as competitive in the large scale manufacturing of these products compared to oversea economies. 3D printers are bound to change all of that, shifting the manufacturing landscape from the production of many copies of identical objects to a market for unique, personalized designs.
3D printing requires no setup for producing different objects, so personalization comes at no cost. More importantly, printing time and cost are not related to surface complexity, allowing the production of highly detailed objects, with details within the 0.1 millimeter range. Multimaterial printers and hydrographics can combine materials of different optical and mechanical properties on the same object, allowing the fabrication of objects whose properties change at each surface location.
While printing hardware is evolving rapidly, design software is not taking advantage of these new capabilities, in particular the ability to mix materials at a sub-millimeter scale. Throughout this project, we will investigate scalable algorithms to design surfaces with patterns of predictable appearance and mechanical properties. We will validate these algorithms with physical measurements, and provide an open source implementation of our findings. We believe that this will change manufacturing entirely since the inherent advantages of 3D printing hardware will now be available to designers with our software. In particular:
- [WP1] we will compute the optimal mix of materials that approximates desired optical properties;
** [WP2] we will compute the optimal surface microstructure that approximates desired mechanical properties;
- [WP3] we will model structured patterns on surfaces for personalization;
- [WP4] we will transfer patterns to surfaces with computational hydrographics;
- [WP5] we will validate the printed objects by physical measurements using light domes;
- [WP6] we will provide an implementation ready for use.
In our prior work, we have already shown the feasibility of the various sub-problems with current printing hardware. But the algorithms we used are not practical, as they may take long time to compute low resolution surfaces. This project will improve significantly over the state of the art in two main areas. First, we propose to derive scalable algorithms for the solution of the above problems to reach the surface complexity that today's printers are capable of. Second, we will solve all these problems at once, since on an object all surface properties are printed together, so providing separate solutions has no practical impact. For this reason, we have assembled a large team of researchers with expertise in different areas. We are confident that we can address all remaining practical software issues for designers to take full advantage of the hardware they already have.