Abstract

In this research, an integrated KD-Tree cutter size search model is proposed to quickly determine the largest cutters and their accessible surface regions for 3-axis finish machining, without gouging and interference. By using these cutters, the highest material removal rate can be achieved, while maintaining the quality of the machined part. To overcome the problems of existing methods, such as long computational time and low accuracy, this model integrates the vertex KD-tree method with local gouging detection method. In this work, an imaginary cutter model is used to define a cutter, given a cutter contact point (CC point) and a testing point. All the testing points needed are derived from the part STL model. A simple and efficient algorithm is suggested to get rid of the redundant vertices in the STL file of the part. The local gouging detection method identifies the maximum local gouging-free cutter for a CC point as the initial cutter, and this cutter is used to determine the area of the cutter shadow. The cutter shadow is used to define a search range of the testing points to avoid any unnecessary search. Then the vertices covered by the shadow are quickly located by the KD-Tree algorithm from all the vertices, and used as test points to determine the final gouging and interference-free cutter size. The proposed model is tested with a hairdryer mould example. The results show that the model is not only computationally efficient, but also highly accurate. In addition, the model is suitable for any types of milling cutters, and ready to implement in the CAD/CAM software. Keywords: STL File, CATIA, Cutter Size Selection, KD-Tree search, Boundary, Gouging, Interference, Accuracy, and Computational Time.