科研之友
微信
新浪微博
Facebook
分享链接
摘要
The surface performances of aerospace materials machined by micro-EDM directly affect the reliability of aeronautical components. However, there are two main problems that must be solved by research in this field. The first is that even if all the machining parameters are kept the same, different materials, pulse generators, and micro-EDM machining types also affect the surface performance. The second problem is that traditional surface evaluation parameters cannot reflect the actual situation accurately. However, research on the surface performance is always a small part of a larger study, and usually not systematic enough. In this study, a systematic investigation of surface performance was conducted using a comprehensive method. A novel mathematical model that combines support vector machine with a multi-objective genetic algorithm was established. Three materials, two pulse generators, three machining types, and various machining parameters were used as inputs to this new model. From this, new evaluation parameters of surface performance, such as the fractal dimension, recast layer thickness, and surface hardness, were generated to use as output parameters. Afterwards, relevant experiments that matched the model input parameters were conducted for comparison. Based on the comprehensive method, the comparative results indicated that the errors between the predicted and experimental values were less than 7%. The developed mechanism based on the predicted and experimental results is discussed in depth in this report, and suggestions on how to utilize this information to machine components with improved surface performance are proposed.
- 出版日期2018-5
- 单位厦门大学
