Data structures and algorithms 3: multi-dimensional searching and computational geometry
Data structures and algorithms 3: multi-dimensional searching and computational geometry
Model referenced monitoring and diagnosis - application to the manufacturing system
Computers in Industry
Real-time shaded NC milling display
SIGGRAPH '86 Proceedings of the 13th annual conference on Computer graphics and interactive techniques
An Algorithm for Geometric Set Operations Using Cellular Subdivision Techniques
IEEE Computer Graphics and Applications
An optimal algorithm for finding segments intersections
Proceedings of the eleventh annual symposium on Computational geometry
3d Modeling Using the Acis Kernel and Toolkit
3d Modeling Using the Acis Kernel and Toolkit
Methods for Detecting Errors in Numerically Controlled Machining of Sculptured Surfaces
IEEE Computer Graphics and Applications
Techniques for accelerating B-rep based parallel machining simulation
Computer-Aided Design
| Hi-index | 0.00 |
Continued progress in the area of solid modeler based machining process simulation is hindered by the complexity growth that occurs for a large number of tool paths n. For this reason, many researchers have adopted the Z-buffer approach. Boundary-representation (B-rep), however, remains the dominant choice for commercial modelers. This paper begins by reviewing the current state of solid modeler based machining simulation. Using an industrial example, the growth rate, for a simple feed rate scheduling application, is estimated to be O(n1.5). It is shown that round robin parallel scheduling quickly becomes inefficient due to the fraction of time spent on tool swept volume Boolean subtractions. The tool path sequence is next heuristically subdivided into nearly equal size neighbor groups. Only the Boolean subtractions required for accurate simulation are included in the group. Each group is then simulated in parallel, achieving a greatly reduced wall clock running time. Computational geometry methods are described that permit rapid identification of tool path neighbors. It is shown that, under practical assumptions, the total number of tool path neighbor pairs is O(n), justifying the benefit of parallel processing. Both dual CPU and networked parallel solutions are implemented. Geometric images and running time plots are included to illustrate. Discussion is included, with proposed steps to further reduce calculation time.