Technology And Neural Network Forecasting System Of Friction Stir Welding Of Magnesium Alloys

Friction stir welding (FSW) is a solid state joining process which was invited by The Welding Institute in 1991. It has been protected by the worldwide patent. Magnesium and magnesium alloys possess very low density, high specific strength and stiffness and good dimension stability capabilities. So the connection of magnesium alloys will be a focus. When the magnesium alloys are welded by TIG or MIG, the metal in the seam will be oxidized acutely. Some defects such as cracks, voids and shrinkages appear. But the metal doesn't melt in the process of FSW. So it's potential to connect magnesium alloys by FSW. At the present time, most of the researches focus on the FSW of aluminum alloys while few researches focus on the FSW of magnesium alloys. The study provides some theoretical references to research the FSW technology of magnesium alloys.The paper designs the probe which is used to weld magnesium AZ31B plates. The material and structure of the probe are optimized by the experiments. It also studies the influence on the quality of the FSW joint caused by rotation speed, welding speed and pressure. The range of the parameters is gotten. Through metallographic experiments, analysis of hardness and tensile experiments, it can be decided that FSW is an effective way to connect magnesium alloys. Through the tensile experiments of high temperature, it's gotten that the structure of the FSW seam hasn't superplasticity. The significance sequence of processing parameters of friction stir welding has been calculated by SN plan design, and the optimized parameters are gotten. The paper establishes the neural network and the user interfaces. It successfully forecasts the strength and the percentage elongation. The forecasted value can satisfy the demand of the projects. The network provides a way to evaluate and forecast the strength of the FSW joints.