Research On 3-Dimensional Intellignet Processing CAM For High Power CO₂ Laser

Research On 3-Dimensional Intelligent Processing CAM For High Power CO2 Laser ABSTRACT This paper discribes the whole process of a 3D intelligent processing CAM system for high power CO2 laser. Author researches the movement of five-axis flying optics laser machine and its characteristic. The laser beam is propagated from laser output window to workpiece surface through laser head by the mirrors and lens. On workpiece surface a certain of laser spot moves along processed trajectory on 3D workpiece. This 3D spatial movement of laser head is basic to conduct 3D laser processing. However, only 3D spatial movement is not enough for 3D laser processing. The laser parameters and processing paramters is key factors to influence final processing result. In this paper above two aspects have been mainly researched: 3D processing trajectory and 3D processing parameters. The intelligence of 3D laser processing shows in the creation of 3D intelligent trajectory and processing parameters self adaption according to trajectory. First of all the characteristic of laser 3D intelligent processing for high power CO2 laser has been analysed. As the flying optics must be used in 3D laser processing, the laser propagation and focusing process are effected by many factors and finally influence processing quality , it is important to make it clear how these factors influence results on five-axis laser machine when the research on 3D intelligent laser processing CAM begins for this research is based on the platform of five-axis laser machine. The laser head must keep vertical to workpiece surface during laser processing. This requirement makes the off-line auto programming difficult in 3D laser CAM system because of complicated algorithm and hard programming. By deeply analysis of data structure of CAD system, the method to extract 3D spatial trajectory data has been found , a series of algorithm of how to get data from CAD file have been put iii I眫{~t~ . m forward in the developed CAM system. For the first time Hopfield neural network is used to plan trajectory for 3D laser processing. This has been proved to be an effective approach to intelligent optimization for laser trajectory planning. With this method not only trajectory data from CAD model can be easily generated, but also boundary and hole edge data of 3D workpiece can be obtained automatically. This intelligent algorithm reduced programming time greatly. Simulation and collision check are important parts in CAD/CAM system.In this paper many algorithms about simulation and collision check are discussed. A quick and less memory consumeing algorithm was adopted for 3D laser CAM system. Laser parameters and processing parameters have been investigated in 3D laser cutting. Especially the influence of incident angle on cutting quality was paid more attention when the laser beam could not be vertical to workpiece surface. Artifical neural network has been built up to analyze the incident angle and movement on cutting result quantitatively. Furthermore expert system for laser cutting parameters optimization and prediction based on test data is set up by BP network self study. Then the parameters subprogram for laser cutting can be created automatically.