تجزیه عملیات برای ویژگی های سطح فرم آزاد در برنامه ریزی عملیات

Machining operations for freeform surface features usually include roughing, finishing for the bottom surface, tapering, and corner rounding. Strategies and algorithms to decompose the overall task into these operations are presented. The decomposition aims at minimizing machining time within the constraints of the specified surface roughness and tolerance, and machine tool safety. The roughing operation can be further decomposed into sub-operations for multiple tools. There are two strategies to decompose the finishing operation into sub-operations: one is based on multiple tools and another is based on multiple tool path patterns. The approaches to select the optimal decomposition values (tool diameters, surface slopes) that minimize machining time are presented. These algorithms are being integrated into a rapid-prototyping service for web-based machining. Design for manufacturability and maximizing process automation are the key priorities in process planning and operation decomposition.

Freeform surfaces are often used in the designs of electronic appliances, automobiles, and airplanes. They allow designers describe part geometries with functional and aesthetic goals in a convenient manner. In the past decade, freeform surface design and machining has attracted much attention from both academic researchers and industry practitioners. Recent research work in this field has mostly focused on feature recognition and tool path planning. Feature-based CAD/CAM [5], [8], [9], [12] and [16] also aims to reduce the interactions between designers and fabricators and to automate process planning. In feature based milling, a part is described as a starting volume or stock and a set of features, representing the volumes removed by machining. Integration of manufacturing knowledge into the design tool allow to check the manufacturability of each feature at the design stage. Thus feature-based CAD can reduce the complexity of operation decomposition, process planning time, and minimize the number of design/fabricator iterations. The purpose of tool path planning is to generate gouge-free tool paths for a desired geometry, with a certain tool and given cutting parameters. In order to meet given requirements on surface finish while minimizing machining time, the tool path pattern should cover the surface as uniformly as possibly. Recently, much research concerns such algorithms to define optimal tool paths for various tools, part geometries, and machines [1], [2], [3], [4], [7] and [13]. Most commercial CAD/CAM systems have the capability to design and machine some kinds of freeform surface features. However, process planning and operation decomposition still are major challenges. Some of the problems encountered include: • Freeform surface features are often machined in multiple operations: roughing, finishing, tapering, and corner rounding. These features have to be decomposed into more cohesive, lower level operations before they can be processed by the tool path generator. • Tool and cutting parameter selections are critical to both machining quality and cutting efficiency. A large tool may yield high cutting efficiency but cannot clean up corners in a freeform feature, while a small tool can cover all areas, but at a low cutting efficiency. A combination of tools of different sizes is thus typically used. During process planning, the tool and cutting parameters for each operation must be selected to meet design specifications (roughness and tolerance) and maintain machine tool safety, while minimizing machining time. • Not all manufacturability problems can been detected at the design stage. Some are related to the chosen tools and their operational parameters, which are only selected at the process planning stage. For example, designers often design pockets that are narrower and deeper than what can be reached with available tools. Operation decomposition builds a linkage between the feature recognizer and the tool path planner. Its separate functions include: • Decomposing a feature into a list of machinable operations (roughing, finishing, tapering, and corner rounding) • Decomposing the roughing operation into sub-operations for multiple tools • Decomposing the finishing operation into sub-operations for multiple tools and multiple tool path patterns • Selecting the optimal decomposition values (tool diameters, surface slopes) that minimize overall machining time After operation decomposition, a list of individual operations each with their local geometries, chosen tools, and applicable type of tool path patterns are forwarded to the tool path planner. The prime concerns are full automation and (almost) guaranteed manufacturability in order to reduce the interaction between designers and manufacturers. In this paper, strategies and algorithms to automatically decompose a freeform surface feature into a list of feasible operations are presented. The features considered are 3-axis milling features, accessible from a predefined direction. They include 2.5D extruded pocket geometries which may have a freeform surface.

Machining operations for freeform surface features include roughing, finishing for the bottom surface, tapering for the wall, and corner rounding for the top rim. In this paper, strategies and algorithms to decompose the overall task into these operations and to further decompose the latter into sub-operations are presented. The decomposition aims at minimizing machining time within the constraints of the specified surface roughness and tolerance, and machine tool safety. These algorithms are being integrated into CyberCut [10], [11] and [15] service for web-based machining at U. C. Berkeley. Design for manufacturability and fully automated process planning are the prime concerns in our environment, and these concerns are also reflected in the operation decomposition process. There are two main limitations in our current work: (1) In the optimal tool selection based on the decomposition maps, the complexity of exhaustive searching exponentially increases with the number of tools. In practice, a reasonable way is to limit the number of selected tools to 2 or 3, but the selected set of tools may not be the global optimal point though not far from it; (2) In the finishing operation decomposition for multiple tools, the surface offset and self-intersection detection algorithms usually is time-consuming. These problems will be our future research focuses.1