Abstract
Processing CAD models of large sized, structurally complex, and dominantly sculptured objects for free-form thick-layered object manufacturing (FF-TLOM) is a non-trivial task. Algorithmic support is needed, especially when the applicators of the FF-TLOM technology are not specialized experts. The research team of the authors has developed and programmed a system of such algorithms to prepare NURBS-based assembly models for layered manufacturing, but also to control both the cutting tool and the manufacturing equipment. This paper starts with a brief overview of the FF-TLOM technology and the concept of reducing a CAD model to fabrication elements. Then, it explains the high-level concepts supporting the algorithms, and presents the details of their operation. The logically linked system of algorithms is applied to find the best position for segmentation. The best segmentation orientation can be found in a computation-intensive, multi-criterion optimization process. It is difficult to simplify and speed up, for the valuation of the parameters of the objective function cannot be done without a preliminary computation of the segments and layers.