Study On Welding Process And Process Optimization Model Of TA5 Titanium Alloy TIG Welding

National defenses have been strongly developing in our country and the advanced material of titanium alloy has been much applied in naval ships building.However,the welding technology falls terribly behind such countries as US,Russia,Japan,etc.Considering the situation,our country is trying his best to catch up and break through kinds of technology challenges.During the welding of titanium alloy,the mechanical property of welding joint is the key factor to the quality of key parts.Thus,it is necessary to predict the mechanical property of welding joint and inspect the safety in order to guarantee the quality of workpieces.This paper,by carrying out a series of welding tests and mechanical property tests and combining with the method of artificial neural network(ANN),studies the TIG welding joint of TA5 titanium alloy applied in the outfitting workpieces and establishes a prediction model of tensile properties of TIG welding joint of TA5 titanium alloy based on ANN.The welding process of TIG welding is relatively stable,and the weld formed by this welding method has a better forming effect.In the TIG welding process of titanium alloy,the process parameters such as welding current,arc voltage,welding speed,and flow rate of shielding gas will have certain influence on the appearance of the weld and the microstructure of the weld,which will also affect Mechanical properties of the joint.Within the given welding process parameters,this paper observes the surface morphology of welds under different welding parameters,and measures the penetration and width of the weld under different parameters,and then obtains the aspect ratio of the weld,and then selects Optimal parameters.The results show that with the increase of current,the weld penetration,melting width and heat-affected zone width increase.With the increase of velocity,the weld penetration,melting width and heat-affected zone width are reduced,and the welding current is in the welding process.There is a matching value with the welding speed,and the aspect ratio of the weld should be optimal.In the process scope of this paper,the best process parameters of the 5mm TA5 titanium alloy during TIG welding are: welding current is 110 A,welding speed is 150mm/min.In addition,the flow rate of the welding torch and the back protective gas also needs to be controlled.In terms of performance,as the welding current increases,the tensile strength of the tensile test piece becomes smaller.As the welding speed increases,the tensile strength increases continuously,which is mainly related to the welding heat input.In this paper,the mechanical properties of TA5 welded joints are tested.After the test,we obtained relatively comprehensive welding conditions and data records related to tensile properties,and then established a prediction model.This model is mainly based on ANN theory in TA5.Tensile properties of this welded joint after titanium alloy TIG welding.After the dimension reduction and the optimization of the sample space samples in the system,the optimization process and retraining are performed,after which we can better fit the current data,and the fit has Higher prediction accuracy.The artificial neural network method is used to establish the welding process parameter optimization model,which will effectively avoid the initial exploration time and cost of a large number of process parameters,and can save a lot of money for the enterprise,thus improving the production efficiency of the enterprise,and the workers in the enterprise also have Higher production enthusiasm has better practical application value in production and construction.Through the above research,it can effectively avoid the large amount of manpower and material resources consumed by enterprises,improve the production efficiency of enterprises and the enthusiasm of workers,and have high engineering application value.A prediction model of tensile properties of TIG welding joint of TA5 titanium alloy is established based on the method of ANN,which,after the dimensionality reduction and the optimizing process of the sample space,can well achieve the test measurement result and has very high prediction accuracy.The above study can effectively avoid wasting a good deal of manpower and materials and improve the production efficiency and the workers' enthusiasm,which has a high engineering application value.