Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
Milling is a most common type of material removal process by rotating tools to create a variety of features on a part. The material is removed in a controlled way by sweeping a rotating cylindrical tool along the specific trajectory known as milling tool path. For any specified shape and volume of material to be removed, there are a variety of possible tool paths and cutting conditions. These different possibilities can be evaluated based on various geometric as well as process related factors. Milling time is one of the most important factors in evaluating the process efficiency, which depends upon the tool path planning and the behavior of the physical machine tool. Tool path planning usually considers factors such as limitations of cutting tools, tolerance requirement and workpiece geometry etc., which directly influence the milling process time. Further, as milling process involves physical interaction between tool and workpiece, the wear and tear of milling tools is an important issue. As the breakdown of cutting tools is detrimental for productivity; the tool paths must be evaluated based on other important physical considerations like cutting forces and chatter. A number of cutting force models on the preselected cutting parameters with high reliable prediction capability are already available in the literature, however, a simplified category of constant engagement zones are usually assumed to exist. Engagement zone may vary along the tool path as it depends on instantaneous in-process workpiece geometry and hence imparting a change in the cutting forces also. Thus one of the objectives of this work is to present a system to verify a milling process plan to incorporate arbitrary tool paths and in-process changes in workpiece geometry. Among the available tool paths, contour parallel tool paths are the most widely used tool paths for 2D milling operations. A number of exact as well as approximate methods are available for offsetting a closed boundary in order to generate a contour parallel tool path; however, most of methods are inherently incapable of dealing with complex problems (change in topology and self intersecting feature) during offsetting and require highly efficient computational routines to identify and rectify these problems. In this work, a boundary value formulation of the offsetting problem is studied and a fast marching method based solution for tool path generation is presented. This method handles the topological changes during offsetting naturally and deals with the generation of discontinuities in the slopes by including an "entropy condition" in its numerical implementation. A number of examples are presented and computational issues are discussed for tool path generation. Although, the tool path generation methods discussed earlier guarantee to generate a geometrically feasible tool path the in-process engagement is still not constant or its variation is not minimized. This leads to variation of actual radial depth of cut especially at the sharp corners or high curvature profiles, the usual problems encountered due to this variation are (i) left over material at corners (ii) sudden increment in cutting forces. These conditions force the process planner to add more tool passes to the original tool paths or adhere to conservative feed values respectively. In either case, it renders the process plan inefficient. This work presents a method based on signed distance function to generate spiral-out contour parallel tool path generation. This proposed method and algorithms avoids the leftover material at the corners and minimizes the variation of radial depth of cut at each level of contour milling and consequently maintains the same cutting conditions specified as starting cutting parameters which are favorable for process reliability, part quality and tool life. Finally the second type of contour parallel tool path i.e. spiral-in milling is investigated for milling tool path generation with an aim to generate "efficient" geometrically feasible spiral in tool paths which minimize the variation of the milling process from its steady state while minimizing the curvature of the tool path. Further, the tool path for non-convex geometry are developed which are optimized for the stepover and the engagement with no tool retraction involved, which is highly desirable for high speed milling of arbitrary pockets.
Daniel Kressner, Axel Elie Joseph Séguin, Gianluca Ceruti