Efficient NC Toolpath Generation for 3-axis Machining of 3-Dimensional Freeform Faceted Surfaces Developed from Digital Images

Abstract

In recent years sculptured surfaces have gained a significant importance in engineering as well as in artistic applications. These surfaces are generally machined using Computer Numeric Control (CNC) machines for achieving required surface accuracy and repeatability. For CNC machining of such complex shapes/surfaces one needs accurate toolpath data. Manual computation of such toolpaths is either impossible or it requires a lot of time as well as expertise in relevant area. There is always high probability of error in manual part programing. Thus toolpath generation for the machining of these surfaces is a non-trivial task in itself. Many commercial solutions are available for automatic generation of Numerical Control (NC) toolpaths for machining of sculptured surfaces. The cost of these software are justified for mass production applications and these software packages generally do not match the need of the small scale artisan or craftsmen because of higher costs and complexity involved in using these software. Thus there is always a need of non-expensive solutions for the automatic generation of toolpaths for such cases. Numerous paradigms have been developed by the researchers to aid this cause. One such solution is proposed in this work. In present work an effort has been made to develop a windows based console application with a single page Graphical User Interface (GUI) that can capture freeform surfaces from digital images. Free form surfaces are generated in the form of point cloud and Stereo Lithography (STL) model. This application provides separate options to process normal image and human portrait. It can also generate toolpath to carve the image into wooden, metallic or stone plaques using a CNC machine with a ball end mill cutter. An algorithm is also proposed to make the raster toolpaths more efficient by eliminating the unnecessary tool location points by comparing their z-map for enabling the CNC machine tool to achieve the higher possible feed rates while machining such parts. The developed algorithm generates the 3D point cloud data which can be directly used for NC toolpath data generation or the point cloud data can be processed into faceted/ STL format which can be then used for toolpath data generation using a ball end mill cutter of user defined diameter. The developed algorithm is tested on various surface models developed from digital images and the results are given in chapter 3 & 4. In the later part of this work an improved algorithm for automatic generation of NC toolpath data with user defined scallop height is developed. The new method for constant scallop height control toolpath generation can be used to compute toolpath data for smooth parametric surfaces represented in STL/ faceted format. This algorithm is developed to enhance the machining efficiency by computation of appropriate feed forward and side step values for zigzag toolpath data used for raster machining of the smooth parametric surfaces taken in STL/faceted data format with user defined scallop height as the governing criteria. Thus the new algorithm is capable of generating toolpath data with bidirectional scallop height control. Two different Bezier surface part shapes have been used in faceted format to prove the capability of the algorithm. The developed toolpaths for the one of the Bezier surface part models have also been filtered through a customized cutter location data reduction algorithm to remove the tool location data points having the same z-height as the neighborhood toolpath locations. All the toolpaths developed in present work are verified in virtual machining environment by using virtual 3-axis milling machine model in a simulator called ToolSim [1] available at State Initiated Design Center (SIDC) in Mechanical Department of Thapar University, Patiala. The graphical as well as the simulation results from the above algorithms are found satisfactory and explained in the required details in this thesis work.