Feature Extraction From Tessellated And Sliced Data in Layered Manufacturing

Kamesh Tata Prototype Express Georges Fadel Clemson University When parts are built in layers, the cross sectional area of each layer has to be defined and filled with a pattern of vectors. This filling process is called hatching and the vectors defme the hatch pattern. To accurately reproduce a three dimensional object, key features need to be identified. In particular, top and bottom surfaces, edges. holes and protrusions must be recognized to ensure the slice plane does carry-the critical information required for the build. This paper describes a technique to extract relevant features from a tessellated model to generate a correct sliced representation. Introduction Layered based manufactUring is based on a mathematical slicing operation that generates slices from an original boundary representation model (tessellated or triangulated model). Different processes use different devices to draw the hatch vectors \vhich represent the ro.;eas to be filled or solidified in a slice. For instance, Stereolithographyand Selective Laser Sintering use a laser to draw the vectors, while other processes may use a pattern mask or a binding agent. Many slicing engines exist, in particular, the slicing program of 3D systems has a proven algorithm,Grogan [1990] described an algorithm based on sorting, Chalasani [1992] proposed two algorithms for slicing 3D objects that also rely on a sorting process to identify areas, Vouzelaud and Bagchi [1992} proposed an adaptive technique, Kirschman and Jara-Almonte [1992] investigated parallel slicing, and Dolenc and lY!akela (1993) suggested slicing procedures which considered the identification ofsome features, notably peaks and flat areas. This paper considers the issue of features and their extraction from the tessellated data. It highlights the capabilities of such an algorithm to identifY manufactUring features and lists the pitfalls and their possible remedys. The STL File The de facto industry STL model representation defines CAD solid models as a set of triangular facets (3D Systems, 1989). Facet models represent solid objects by spatial boundaries which are defined by a set of planar faces. This is a special case of the more general boundary representation which does not require object boundaries to be planar (Mortenson, 1985). In general, the term facet is used to denote any constrained polygonal planar region used to define a model boundary; however, in the SFF community, the term facet is typically understood to mean triangular facet (Rock. 1991). Unfortunately. these facets are stored independently, as ifeach facet was created and tossed into a bucket with no particular ordering and without information relating a given facet to any other facet in the bucket (Rock, 1991).