Slice Overlap-Detection Algorithm for Process Planning in Multiple-Material Stereolithography
Recently, a multiple-material stereolithography (MMSL) system was introduced, which included the novel feature of stacking different photocurable resins to produce a multiple-material part. This process is capable of fabricating intricately detailed parts with smooth surface finish and internal structures of various colors. However, the MMSL system requires a washing process when switching materials, and this additional step increases the fabrication time and also introduces deformations in the build structure as registration errors accumulate. Consequently, reducing the number of material changeovers in this process is a priority, and the minimum number of changeovers is imposed in part by the sweep interference between layers of different materials. Fortunately, the use of low-viscosity resins permits possible fabrication without sweeping. As a result, multiple layers of one material can be continuously built without changing resins, and subsequently, a different low-viscosity material can be added without resulting in sweep obstruction with the original layers. In this paper, a stacking simulation, which determines the maximum number of layers possible to be continuously built between material changeovers, is described. This paper presents a novel interference-detection algorithm, which identifies interference by inspecting the overlap between loop segments. In addition, the algorithm successfully manages the triangulation errors that occur at a curved surface when two materials are adjacent within a layer. Finally, several practical examples are shown to verify the algorithm and provide compelling evidence that the proposed algorithm is effective and applicable in MMSL.
The International Journal of Advanced Manufacturing Technology
Kim, Ho-Chan; Choi, Jae-Won; MacDonald, Eric; and Wicker, Ryan, "Slice Overlap-Detection Algorithm for Process Planning in Multiple-Material Stereolithography" (2010). Mechanical Engineering Faculty Research. 362.