Advanced data exchange for solid freeform fabrication

Solid Freeform Fabrication (SFF) is one of the leading groups of emerging technologies for producing physical objects directly from the computer generated information of the parts. All SFF technologies are enabled by computer hardware and software, from the input geometry to control of the fabrication process. In SFF research, so far, the data formats have been strictly used to store geometric information preferably with minimum size. Current SFF data exchange practice supports only discretized approximations of part geometry. For producing accurate patterns and functional parts, the adequacy of linear approximation is suspect. The most efficient way of minimizing the degradation of geometric entities is to eliminate this intermediate format and directly exchange the source geometry. However, each commercial CAD vendor has its own representation, and choosing a single format is proven somewhat problematic. As an alternative, I have developed a method to approximate a contour from a set of discrete points using high order polynomials and to efficiently generate scan vectors from the high order contour. Our method exploits the so-called algebraic spline (A-spline) curve. The A-spline approximation is capable of representing sliced geometry generated from various types of surfaces, including parametric, implicit, and mixed surface forms. The A-spline approximation method was tested by fabricating freeform objects by Selective Laser Sintering (SLS). The resulting parts showed higher surface quality with none of the tessellation artifacts inherent in part generation from polygonal contours. The run-time efficiency of the method is comparable to that of linear contour processing. SFF methods have demonstrated the potential to manufacture parts from Functionally Gradient Materials (FGM). One of the keys to success is to accurately and systematically represent varying material distribution in the geometry. The dissertation introduces a method called Volumetric Multi-Texturing (VMT) to represent a three dimensional density gradient. The scheme is inspired by volumetric rendering by texturing, which is popular in creating fuzzy objects such as cloud and smoke. By analogy, FGM design is envisaged as the creation of material clouds in a confined geometric space in structured and controllable manner. Another motivation for pursuing this approach is that, based on the survey into expected applications, material gradients will be emphasized near the surface of a part. This method exploits procedural and implicit scheme to design and acquire density information. The implicit procedural approach, as opposed to an input database, allows users to interactively create and modify the design patterns without explicitly changing the individual values in the database. Further, it promises convenience in process planning, and efficiency in data storage and computation time. The material gradient modeler is exercised on a boundary representation (B-Rep) model of the part. Therefore, this scheme can be easily integrated with accessible formats from most of the commercial solid modelers. The theoretical approach, design procedure, and tool path generation for fabrication of example parts are presented in the dissertation.