The Research On Intelligent Design Methods Of Stamping Process And Die For Fuel Tank Cover Based On Knowledge And Its Application

It is highly important to design the stamping process and the die for motorcycle tank cover which is prone to wrinkle and crack.In order to over the inficiency problem of current design practice which depends on engineer's experiences to test dies repeatedly,the intelligent design methods of cold stamping process and dies for motorcycle tank cover are researched by using the numerical simulation technology and KBE(Knowledge Based Engineering)tools.The framework of cold stamping process and die intelligent design system for tank cover(CSPDIDS-TC)is constructed according to the functional requirements of cold stamping process,and the system arrangement is designed which is mainly composed of user interface,application module,decision module,knowledge and data integration module.In order to investigate the drawing defects of the motorcycle fuel tank cover,the forming process of Double-Cover Forming(DCF)was simulated by using DYNAFORM software,and the results show that the cover is prone to two flaws which are sidewall wrinkling and local rupture.Two forming parameters which are draw bead height and blank holding force are analyzed and optimized by using numerical simulation and Orthogonal Experiment Design(OED),and the analysis results indicate that it is effective to suppress the wrinkling defects by increasing the two parameters properly.The optimized part is simulated and experimented,and the results show that the overall forming quality including forming uniformity is improved considerably.Based on the form feature tree of product,the product knowledge integration model(PKIM)is suggested,which combines geometry layer,knowledge layer,feature layer and component layer.The PKIM of tank cover is generated by inheriting and mapping form feature.A knowledge representation method is proposed by using object-oriented method and production rules,and the credibility and threshold is incorporated into the knowledge model to realize decision-making based on experience knowledge.The PKIM is implemented by using UG/KF language intent!.The direct and indirect knowledge discovery method based on numerical simulation are analyzed.A direct method of knowledge discovery based on the causal logic is proposed and a knowledge discovery instance is given about the relationships between side wall wrinkling defect and blank holder force and drawbead parameters selected empirically by engineers.The indirect knowledge discovery method based on data reduction is explained,which includes rough set reduction,principal component analysis and decision-tree reduction,and implicit knowledge is acquired about the relationship between five parameters and the forming quality index extract from the simulation data of drawing process.The reasoning mechanism of the CSPDIDS-TC is designed which is composed of case and rule knowledge.The forward and backward reasoning control strategy and conflict resolution mechanism are given,the case based reasoning by using the uncertainty knowledge of credibility factors is proposed,and an instance by using the uncertainty matching and conflict resolution is given.The intelligent design system interface and the knowledge module using UG/Open and UG/KF tools were designed and the intelligent design of stamping process and parametric assembly of die for tank cover were implemented.The research results of the knowledge discovery methods based on numerical simulation and hierarchical PKIM are beneficial to the implicit knowledge acquisition and formal knowledge management of the complicated cold stamping process design.