.5 A06 Of Aluminum Alloy Electron Beam Welded Alloy Elements Burning Behavioral Research

5A06 aluminium alloy is widely used in industry whose strengthen mainly depends on Mg element. Part of Mg's evaporation loss occurs during electron beam welding (EBW), which inevitably deteriorates mechanical properties of welded joint. In this thesis, Mg's evaporation loss behavior and its influence on mechanical properties of welded joint, and influence of welding parameters on evaporation loss behavior were investigated. Furthermore the measures to reduce evaporation loss and improve mechanical properties of welded joint were put forward. Intelligent optimization of welding parameters was conducted by using artificial neural network (ANN).Distribution of Mg in melt pool and its effects on mechanical properties of welded joint were investigated by EDX analysis, microhardness test and tensile properties test. The results show that evaporation loss rate in the center of melt pool is higher than that of margin and the evaporation loss rate in upside is higher than that of downside; the trend of microhardness in melt pool is agreed with that of Mg content, namely, microhardness in central region is lower than that of margin and microhardness in upside lower than that of downside. Comparisons between experimental results of different samples show that microhardness and tensile strength fall while evaporation loss rate of Mg content increases.Influence of welding parameters on fusion penetration and evaporation loss, and its mechanism were investigated by orthogonal experiments and statistical analysis. The results show that accelerating voltage, beam current and welding speed are the main welding parameters which influence fusion penetration and evaporation loss; fusion penetration increases when accelerating voltage, beam current increase and welding speed minishes; evaporation loss rate reduces when accelerating voltage, beam current and welding speed increase. Consequently, evaporation loss of Mg element can be controlled in a reasonable range with insured fusion penetration by proper increasing of the three parameters.Fusion penetration and evaporation loss rate is predicted based on ANN. A BP(Back Propagation) neural network mapping model from 'welding parameters space' to 'fusion penetration and evaporation loss rate' is established. The trained network can predict fusion penetration and evaporation loss rate in a definite range through graphical user interfaces, which has good verifying precision. The network offers an effective way to assist to optimize welding parameters and control evaporation loss.