Research On Axial-force Control Of 6061-T6 Aluminum Alloy Sheet In HR-FSW

In friction stir welding,the welding quality will cause serious defects such as deformation and effective thickness reduction due to welding parameters and some uncertainties,which will make the joint structure invalid.To solve this problem,a complete set of friction stir welding axial force detection and control system is designed in this paper.The change trend of the axial force and the effect of the axial force on the structure and mechanical properties of the welded joint of the 6061-T6 aluminum alloy thin plate with high speed friction stir welding joint was studied.The joint forming mechanism is analyzed.And the process parameters of 1mm6061-T6 aluminum alloy sheet high speed friction stir welding is improved.Finally,based on the fuzzy control algorithm,the axial force closed-loop control system was built.In this paper,orthogonal test is used to study the influence of different welding parameters on axial force and welding quality.Experimental research shows that the welding speed has the greatest influence on the welding axial force.The fitted curve between the welding speed and the welding axial force is:y=-1.5e-06*x²+0.0017*x+0.58,where y is the welding axial force in k N;x is the welding speed in mm·min^-1.At the same time,the best welding parameters of 1mm thick 6061-T6aluminum alloy sheet were obtained.The welding speed was 200mm·min^-1,the spindle speed was 10000rpm,and the shaft shoulder pressure was 0.06mm.Through the combination of mechanical properties analysis and structure observation on the welded joints,the welding quality under different axial forces is studied.The research shows that when the welding axial force is 0.9k N?1k N,a welded joint with excellent performance and beautiful appearance can be obtained.At this time,the tensile strength of the joint can reach 201MPa and account for 70%of the strength of the base material.The fracture rules have obvious necks.Shrinking phenomenon.The hardness value of each area of the joint is evenly distributed without obvious mutation.The nucleus area is dominated by small equiaxed crystals,the organization of the heat-affected zone is uniformly transitioned,and the narrower heat-affected zone has less effect on the weld seam,and no defects such as tunnel grooves are generated.If the welding axial force is too large or too small,it will cause corresponding welding defects in the joint and reduce the mechanical properties of the joint.Through the BP neural network algorithm,a 3-input and 1-output axial force neural network prediction model is designed.The model has a three-layer structure including:input layer,hidden layer,and output layer.The model uses the Levenberg-Marquardt BP training function(Trainlm),the transfer function between each layer is the unipolar S function logsig,and combined with the actual experimental data to achieve the establishment of the axial force prediction model.The model verification shows that the model prediction results have concentrated error,small difference and high stability.Therefore,the model can effectively predict the welding axial force under the corresponding welding parameters,and the simulation operation effect is good.Taking the axial force as the input of the fuzzy controller and the welding speed as the control object,the fuzzy controller is algorithmically designed to adjust the axial force,and combined with the BP neural network prediction model,a simulation test is carried out.The simulation results show that the axial force closed-loop control system has good stability and strong target value tracking.Finally,the actual hardware connection is realized,and the communication between each unit module is realized.The verification research shows that the welding axial force floats up and down around the target value,the fluctuation range is small,and the system stability is good.The structure of the welded joint is even and dense,and the transition between the areas is uniform,without obvious welding defects.The tensile strength of 195Mpa accounts for about 70%of the base material(290MPa).The hardness distribution is uniform,without obvious mutation,and the mechanical properties of the joint are excellent.In summary,the control system can achieve the control purpose,and can meet the experimental design requirements.