Study On Friction Stir Welding Technics Of Al Cu And Optimization Based On Artificial Neural Networks

Friction stir welding(FSW) is a newly developed technology. It not only has the advantage of traditional friction welding but also breaks the limitation of traditional friction welding which only can weld cylindrical member. Not only straight weld seam but also various kinds of joints in different position can be welded by FSW. From the beginning of the 21st century, it is paid much attention in the welding field. Some industrial nations have used in the engineering. However, FSW is the patent technology of TWI. In the related papers, the content about parameter of FSW and friction stir tools is scarce. In addition, traditional quality testing methods destroy samples, waste material and prolong production time, which cannot satisfy the need of engineering. As known to us all, Al and Cu is widely used in production. Therefore, in order to catch up with the advanced technological level, it is imminent and difficult to use new FSW technology in the joint of Al plate and Cu plate, realize the optimization of FSW parameter and estimation of joint quality.In this paper, design of friction stir welding tool, performance of friction stir welding joints, friction stir welding parameter, SN plan and analysis, quality estimation of joints based on artificial neural networks(ANN) are particularly discussed. The regulation of friction stir tool, performance of friction stir welding joints, friction stir weld parameter is obtained. And ANN technology is used in joint quality estimation of friction stir welding in this paper. Training data of ANN is acquired through cross plan, ANN and user interface are established by MATLAB. After training, the trained network is used to estimate joint quality and to judge whether the joint is good or not. The estimation data is satisfying to the requirement of the engineering, which provides a new way to the estimation of friction stir welding joint quality and optimization of parameter.The results of the research can be used in parameter optimization of FSW and accelerate the application of FSW in colored metal in our nation. Particularly, ANN is first used in the research of optimization of friction stir welding parameter which makes full use of some tests and reduces tedious work, abridges time, save cost and is in favor of the development of friction stir welding technology.