Gouge detection algorithms for sculptured surface NC generation

Two algorithms are presented to detect and correct milling errors which can occur in commonly applied three-axis NC generation procedures for sculptured surface parts. The first method exploits surface normals and the geometry of the tool to accurately characterize the chordal deviation resulting from an actual tool motion. The second technique considers the global nature of an entire tool pass to detect gouging in areas of high curvature variation. This method is capable of detecting gouges in both locally concave and convex regions of curvature. The algorithms are designed in a modular fashion and can be incorporated as part of the general generative procedure. Results of such an implementation are presented which demonstrate the benefits of the gouge detection algorithms.