Font Size: a A A

Investigation On Laser-MIG Hybrid Welding Of TC4 Alloy By Neural Network Optimization

Posted on:2021-11-14Degree:MasterType:Thesis
Country:ChinaCandidate:W HuangFull Text:PDF
GTID:2481306479463994Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
TC4 titanium alloy has excellent properties such as high specific strength,good heat resistance and corrosion resistance,etc.It becomes the key structural material in aerospace field,and it is also widely used in energy,chemical industry and medical fields.In the industrial application,TC4 alloy is often used as welded structure.Consequently,the weldability of TC4 alloy is always paid more attention by people.TC4 titanium alloy has the characteristics of low thermal conductivity and high chemical activity at high temperature.When it is welded by using traditional fusion welding processes,the coarser grains and embrittlement easily occur to weldment,which results in the decrease of joint performance.Compared with traditional fusion welding processes,laser beam welding(LBW)has the advantages of high energy density,large depth-to-width ratio of weld and narrow heat-affected zone(HAZ),etc,which is suitable for the welding of titanium alloy.Laser-MIG hybrid welding developed from LBW has both the deep penetration of LBW and good adaptability to assembly gap of MIG arc welding.Consequently,it has great significance to investigate the laser-MIG hybrid welding of TC4alloy.During welding,welding parameters have great effect on the size of weldment and the performance of welded joint.If the welding parameters are determined improperly,welding defects such as incomplete penetration,burning through and concavity are easily generated in welded joint,which causes the decrease of joint mechanical property.In this paper,the software MATLAB is used to establish the BP neural network and the genetic-algorithm BP neural network(GABP).TC4titanium alloy is taken as the research object,and the exploratory study of this alloy by using laser-MIG hybrid welding is carried out.The mapping relationship among laser power,welding speed,welding current,width of weld and depth of fusion is established by using neural network,thus the appropriate laser-MIG hybrid welding parameters are determined for the TC4 alloy with the thicknesses of 3.5mm and 6mm respectively.Through welding experiments,the effect of LBW and laser-MIG hybrid welding on the microstructure and mechanical properties of welded joint is investigated.By comparision the microstructure and mechanical properties of the two welded joints,the feasibility of LBW and laser-MIG hybrid welding to TC4 titanium alloy with different thicknesses is discussed.The neural network simulation results show that the prediction error rate by the BP neural network models is at the range of 3%-4%.After the optimization by genetic algorithms,the prediction error rate by the BP neural network model is reduced to 1%-3%.According to the calculated results by neural network model,the laser-MIG hybrid welding parameters for TC4 alloy with the thickness of3.5mm are determined as follows:laser power 3.0-4.0k W,welding speed 1.5-2.0m·min-1,and MIG arc welding current about 80A.The laser-MIG hybrid welding parameters for TC4 alloy with the thickness of 6mm are determined as follows:Laser power 3.5-4.0k W,welding speed 0.8-1.2m·min-1,and MIG arc welding current 80-120A.Analyses and tests are carried out to the welded joints after welding.The results show that the microstructure of weldment of TC4 alloy is composed of??martensite.The microstructure of HAZ is changed with the different distance away from weld center.The near-HAZ(near weld metal)is composed of?and??two-phase microstructure,the far-HAZ(near base metal)is composed of?and?two-phase microstructure,and the transition zone(between them)is composed of?,??and?three-phase microstructure.Grains in weldment by LBW are relatively fine,while grains in weldment by laser-MIG hybrid welding are relatively coarse and the size of martensite is larger.Tests of mechanical properties for joints indicate that the microhardness of the two joints in fusion zone(FZ)is the highest,the microhardness in HAZ is decreased to a certain extent,and the microhardness in base metal(BM)is the lowest.Under the condition of appropriate welding parameters,the tensile strength of the two welded joints is higher,which is close to that of the BM.There are many dimples with the reticular morphology on the tensile fracture of welded joints,and it obviously presents the characteristics of ductile fracture.Under the experimental condition in this paper,LBW is suitable for the welding of TC4 alloy sheet,and the joints with fine grains and good performance can be obtained.Laser-MIG hybrid welding is suitable for the welding of TC4 alloy with medium thickness,and the gas pores in weldment is greatly reduced,and the mechanical performance of welded joint can be effectively improved.
Keywords/Search Tags:TC4 titanium alloy, laser beam welding, laser-MIG hybrid welding, neural network, microstructure, mechanical properties
PDF Full Text Request
Related items