Study On Fuzzy Clustering Approach In Analyzing Defects Of Cold Extruded Part (Id : 6-020)
Gao Jun,Ji Tingwei,Zhao Guoqun,Zhang Chengrui,Yu Dingting
2006-01-01
Abstract:Cold extrusion process, which is widely used in war industry, aviation, automobile, and etc, plays great important roles in the field of metal forming. By analyzing the defects of the cold extruded parts in advance, the problems in the process of cold extrusion process and die designs can be reduced, and the design efficiency can be improved in a certain extent. However, the traditional methods, such as FEM, are subjective and take time. In this paper, fuzzy clustering technique was utilized to analyze the defects of the cold extruded parts. The feature matrix of the cold extruded part was established according to the various attributes of the cold extruded parts and the membership functions, the Max-min algorithm was selected to obtain the fuzzy equivalence relation matrix, and the maximum tree clustering algorithm was used to get the defect types of the cold extruded parts. Finally, the fuzzy clustering approach to analyze the defects of the cold extruded parts was validated through typical cup-shaped cold extruded samples.